lbcd/blockchain/chainio.go

1317 lines
47 KiB
Go
Raw Normal View History

// Copyright (c) 2015-2017 The btcsuite developers
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Use of this source code is governed by an ISC
// license that can be found in the LICENSE file.
package blockchain
import (
"bytes"
"encoding/binary"
"fmt"
"math/big"
"sort"
"time"
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
"github.com/btcsuite/btcd/chaincfg/chainhash"
"github.com/btcsuite/btcd/database"
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
"github.com/btcsuite/btcd/wire"
"github.com/btcsuite/btcutil"
)
var (
// hashIndexBucketName is the name of the db bucket used to house to the
// block hash -> block height index.
hashIndexBucketName = []byte("hashidx")
// heightIndexBucketName is the name of the db bucket used to house to
// the block height -> block hash index.
heightIndexBucketName = []byte("heightidx")
// chainStateKeyName is the name of the db key used to store the best
// chain state.
chainStateKeyName = []byte("chainstate")
// spendJournalBucketName is the name of the db bucket used to house
// transactions outputs that are spent in each block.
spendJournalBucketName = []byte("spendjournal")
// utxoSetBucketName is the name of the db bucket used to house the
// unspent transaction output set.
utxoSetBucketName = []byte("utxoset")
// byteOrder is the preferred byte order used for serializing numeric
// fields for storage in the database.
byteOrder = binary.LittleEndian
)
// errNotInMainChain signifies that a block hash or height that is not in the
// main chain was requested.
type errNotInMainChain string
// Error implements the error interface.
func (e errNotInMainChain) Error() string {
return string(e)
}
// isNotInMainChainErr returns whether or not the passed error is an
// errNotInMainChain error.
func isNotInMainChainErr(err error) bool {
_, ok := err.(errNotInMainChain)
return ok
}
// errDeserialize signifies that a problem was encountered when deserializing
// data.
type errDeserialize string
// Error implements the error interface.
func (e errDeserialize) Error() string {
return string(e)
}
// isDeserializeErr returns whether or not the passed error is an errDeserialize
// error.
func isDeserializeErr(err error) bool {
_, ok := err.(errDeserialize)
return ok
}
// isDbBucketNotFoundErr returns whether or not the passed error is a
// database.Error with an error code of database.ErrBucketNotFound.
func isDbBucketNotFoundErr(err error) bool {
dbErr, ok := err.(database.Error)
return ok && dbErr.ErrorCode == database.ErrBucketNotFound
}
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// -----------------------------------------------------------------------------
// The transaction spend journal consists of an entry for each block connected
// to the main chain which contains the transaction outputs the block spends
// serialized such that the order is the reverse of the order they were spent.
//
// This is required because reorganizing the chain necessarily entails
// disconnecting blocks to get back to the point of the fork which implies
// unspending all of the transaction outputs that each block previously spent.
// Since the utxo set, by definition, only contains unspent transaction outputs,
// the spent transaction outputs must be resurrected from somewhere. There is
// more than one way this could be done, however this is the most straight
// forward method that does not require having a transaction index and unpruned
// blockchain.
//
// NOTE: This format is NOT self describing. The additional details such as
// the number of entries (transaction inputs) are expected to come from the
// block itself and the utxo set. The rationale in doing this is to save a
// significant amount of space. This is also the reason the spent outputs are
// serialized in the reverse order they are spent because later transactions
// are allowed to spend outputs from earlier ones in the same block.
//
// The serialized format is:
//
// [<header code><version><compressed txout>],...
//
// Field Type Size
// header code VLQ variable
// version VLQ variable
// compressed txout
// compressed amount VLQ variable
// compressed script []byte variable
//
// The serialized header code format is:
// bit 0 - containing transaction is a coinbase
// bits 1-x - height of the block that contains the spent txout
//
// NOTE: The header code and version are only encoded when the spent txout was
// the final unspent output of the containing transaction. Otherwise, the
// header code will be 0 and the version is not serialized at all. This is
// done because that information is only needed when the utxo set no longer
// has it.
//
// Example 1:
// From block 170 in main blockchain.
//
// 1301320511db93e1dcdb8a016b49840f8c53bc1eb68a382e97b1482ecad7b148a6909a5c
// <><><------------------------------------------------------------------>
// | | |
// | version compressed txout
// header code
//
// - header code: 0x13 (coinbase, height 9)
// - transaction version: 1
// - compressed txout 0:
// - 0x32: VLQ-encoded compressed amount for 5000000000 (50 BTC)
// - 0x05: special script type pay-to-pubkey
// - 0x11...5c: x-coordinate of the pubkey
//
// Example 2:
// Adapted from block 100025 in main blockchain.
//
// 0091f20f006edbc6c4d31bae9f1ccc38538a114bf42de65e868b99700186c64700b2fb57eadf61e106a100a7445a8c3f67898841ec
// <><----------------------------------------------><----><><---------------------------------------------->
// | | | | |
// | compressed txout | version compressed txout
// header code header code
//
// - Last spent output:
// - header code: 0x00 (was not the final unspent output for containing tx)
// - transaction version: Nothing since header code is 0
// - compressed txout:
// - 0x91f20f: VLQ-encoded compressed amount for 34405000000 (344.05 BTC)
// - 0x00: special script type pay-to-pubkey-hash
// - 0x6e...86: pubkey hash
// - Second to last spent output:
// - header code: 0x8b9970 (not coinbase, height 100024)
// - transaction version: 1
// - compressed txout:
// - 0x86c647: VLQ-encoded compressed amount for 13761000000 (137.61 BTC)
// - 0x00: special script type pay-to-pubkey-hash
// - 0xb2...ec: pubkey hash
// -----------------------------------------------------------------------------
// spentTxOut contains a spent transaction output and potentially additional
// contextual information such as whether or not it was contained in a coinbase
// transaction, the version of the transaction it was contained in, and which
// block height the containing transaction was included in. As described in
// the comments above, the additional contextual information will only be valid
// when this spent txout is spending the last unspent output of the containing
// transaction.
type spentTxOut struct {
compressed bool // The amount and public key script are compressed.
version int32 // The version of creating tx.
amount int64 // The amount of the output.
pkScript []byte // The public key script for the output.
// These fields are only set when this is spending the final output of
// the creating tx.
height int32 // Height of the the block containing the creating tx.
isCoinBase bool // Whether creating tx is a coinbase.
}
// spentTxOutHeaderCode returns the calculated header code to be used when
// serializing the provided stxo entry.
func spentTxOutHeaderCode(stxo *spentTxOut) uint64 {
// The header code is 0 when there is no height set for the stxo.
if stxo.height == 0 {
return 0
}
// As described in the serialization format comments, the header code
// encodes the height shifted over one bit and the coinbase flag in the
// lowest bit.
headerCode := uint64(stxo.height) << 1
if stxo.isCoinBase {
headerCode |= 0x01
}
return headerCode
}
// spentTxOutSerializeSize returns the number of bytes it would take to
// serialize the passed stxo according to the format described above.
func spentTxOutSerializeSize(stxo *spentTxOut) int {
headerCode := spentTxOutHeaderCode(stxo)
size := serializeSizeVLQ(headerCode)
if headerCode != 0 {
size += serializeSizeVLQ(uint64(stxo.version))
}
return size + compressedTxOutSize(uint64(stxo.amount), stxo.pkScript,
stxo.version, stxo.compressed)
}
// putSpentTxOut serializes the passed stxo according to the format described
// above directly into the passed target byte slice. The target byte slice must
// be at least large enough to handle the number of bytes returned by the
// spentTxOutSerializeSize function or it will panic.
func putSpentTxOut(target []byte, stxo *spentTxOut) int {
headerCode := spentTxOutHeaderCode(stxo)
offset := putVLQ(target, headerCode)
if headerCode != 0 {
offset += putVLQ(target[offset:], uint64(stxo.version))
}
return offset + putCompressedTxOut(target[offset:], uint64(stxo.amount),
stxo.pkScript, stxo.version, stxo.compressed)
}
// decodeSpentTxOut decodes the passed serialized stxo entry, possibly followed
// by other data, into the passed stxo struct. It returns the number of bytes
// read.
//
// Since the serialized stxo entry does not contain the height, version, or
// coinbase flag of the containing transaction when it still has utxos, the
// caller is responsible for passing in the containing transaction version in
// that case. The provided version is ignore when it is serialized as a part of
// the stxo.
//
// An error will be returned if the version is not serialized as a part of the
// stxo and is also not provided to the function.
func decodeSpentTxOut(serialized []byte, stxo *spentTxOut, txVersion int32) (int, error) {
// Ensure there are bytes to decode.
if len(serialized) == 0 {
return 0, errDeserialize("no serialized bytes")
}
// Deserialize the header code.
code, offset := deserializeVLQ(serialized)
if offset >= len(serialized) {
return offset, errDeserialize("unexpected end of data after " +
"header code")
}
// Decode the header code and deserialize the containing transaction
// version if needed.
//
// Bit 0 indicates containing transaction is a coinbase.
// Bits 1-x encode height of containing transaction.
if code != 0 {
version, bytesRead := deserializeVLQ(serialized[offset:])
offset += bytesRead
if offset >= len(serialized) {
return offset, errDeserialize("unexpected end of data " +
"after version")
}
stxo.isCoinBase = code&0x01 != 0
stxo.height = int32(code >> 1)
stxo.version = int32(version)
} else {
// Ensure a tx version was specified if the stxo did not encode
// it. This should never happen unless there is database
// corruption or this function is being called without the
// proper state.
if txVersion == -1 {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return offset, AssertError("decodeSpentTxOut called " +
"without a containing tx version when the " +
"serialized stxo that does not encode the " +
"version")
}
stxo.version = txVersion
}
// Decode the compressed txout.
compAmount, compScript, bytesRead, err := decodeCompressedTxOut(
serialized[offset:], stxo.version)
offset += bytesRead
if err != nil {
return offset, errDeserialize(fmt.Sprintf("unable to decode "+
"txout: %v", err))
}
stxo.amount = int64(compAmount)
stxo.pkScript = compScript
stxo.compressed = true
return offset, nil
}
// deserializeSpendJournalEntry decodes the passed serialized byte slice into a
// slice of spent txouts according to the format described in detail above.
//
// Since the serialization format is not self describing, as noted in the
// format comments, this function also requires the transactions that spend the
// txouts and a utxo view that contains any remaining existing utxos in the
// transactions referenced by the inputs to the passed transasctions.
func deserializeSpendJournalEntry(serialized []byte, txns []*wire.MsgTx, view *UtxoViewpoint) ([]spentTxOut, error) {
// Calculate the total number of stxos.
var numStxos int
for _, tx := range txns {
numStxos += len(tx.TxIn)
}
// When a block has no spent txouts there is nothing to serialize.
if len(serialized) == 0 {
// Ensure the block actually has no stxos. This should never
// happen unless there is database corruption or an empty entry
// erroneously made its way into the database.
if numStxos != 0 {
return nil, AssertError(fmt.Sprintf("mismatched spend "+
"journal serialization - no serialization for "+
"expected %d stxos", numStxos))
}
return nil, nil
}
// Loop backwards through all transactions so everything is read in
// reverse order to match the serialization order.
stxoIdx := numStxos - 1
stxoInFlight := make(map[chainhash.Hash]int)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
offset := 0
stxos := make([]spentTxOut, numStxos)
for txIdx := len(txns) - 1; txIdx > -1; txIdx-- {
tx := txns[txIdx]
// Loop backwards through all of the transaction inputs and read
// the associated stxo.
for txInIdx := len(tx.TxIn) - 1; txInIdx > -1; txInIdx-- {
txIn := tx.TxIn[txInIdx]
stxo := &stxos[stxoIdx]
stxoIdx--
// Get the transaction version for the stxo based on
// whether or not it should be serialized as a part of
// the stxo. Recall that it is only serialized when the
// stxo spends the final utxo of a transaction. Since
// they are deserialized in reverse order, this means
// the first time an entry for a given containing tx is
// encountered that is not already in the utxo view it
// must have been the final spend and thus the extra
// data will be serialized with the stxo. Otherwise,
// the version must be pulled from the utxo entry.
//
// Since the view is not actually modified as the stxos
// are read here and it's possible later entries
// reference earlier ones, an inflight map is maintained
// to detect this case and pull the tx version from the
// entry that contains the version information as just
// described.
txVersion := int32(-1)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
originHash := &txIn.PreviousOutPoint.Hash
entry := view.LookupEntry(originHash)
if entry != nil {
txVersion = entry.Version()
} else if idx, ok := stxoInFlight[*originHash]; ok {
txVersion = stxos[idx].version
} else {
stxoInFlight[*originHash] = stxoIdx + 1
}
n, err := decodeSpentTxOut(serialized[offset:], stxo,
txVersion)
offset += n
if err != nil {
return nil, errDeserialize(fmt.Sprintf("unable "+
"to decode stxo for %v: %v",
txIn.PreviousOutPoint, err))
}
}
}
return stxos, nil
}
// serializeSpendJournalEntry serializes all of the passed spent txouts into a
// single byte slice according to the format described in detail above.
func serializeSpendJournalEntry(stxos []spentTxOut) []byte {
if len(stxos) == 0 {
return nil
}
// Calculate the size needed to serialize the entire journal entry.
var size int
for i := range stxos {
size += spentTxOutSerializeSize(&stxos[i])
}
serialized := make([]byte, size)
// Serialize each individual stxo directly into the slice in reverse
// order one after the other.
var offset int
for i := len(stxos) - 1; i > -1; i-- {
offset += putSpentTxOut(serialized[offset:], &stxos[i])
}
return serialized
}
// dbFetchSpendJournalEntry fetches the spend journal entry for the passed
// block and deserializes it into a slice of spent txout entries. The provided
// view MUST have the utxos referenced by all of the transactions available for
// the passed block since that information is required to reconstruct the spent
// txouts.
func dbFetchSpendJournalEntry(dbTx database.Tx, block *btcutil.Block, view *UtxoViewpoint) ([]spentTxOut, error) {
// Exclude the coinbase transaction since it can't spend anything.
spendBucket := dbTx.Metadata().Bucket(spendJournalBucketName)
serialized := spendBucket.Get(block.Hash()[:])
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
blockTxns := block.MsgBlock().Transactions[1:]
stxos, err := deserializeSpendJournalEntry(serialized, blockTxns, view)
if err != nil {
// Ensure any deserialization errors are returned as database
// corruption errors.
if isDeserializeErr(err) {
return nil, database.Error{
ErrorCode: database.ErrCorruption,
Description: fmt.Sprintf("corrupt spend "+
"information for %v: %v", block.Hash(),
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
err),
}
}
return nil, err
}
return stxos, nil
}
// dbPutSpendJournalEntry uses an existing database transaction to update the
// spend journal entry for the given block hash using the provided slice of
// spent txouts. The spent txouts slice must contain an entry for every txout
// the transactions in the block spend in the order they are spent.
func dbPutSpendJournalEntry(dbTx database.Tx, blockHash *chainhash.Hash, stxos []spentTxOut) error {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
spendBucket := dbTx.Metadata().Bucket(spendJournalBucketName)
serialized := serializeSpendJournalEntry(stxos)
return spendBucket.Put(blockHash[:], serialized)
}
// dbRemoveSpendJournalEntry uses an existing database transaction to remove the
// spend journal entry for the passed block hash.
func dbRemoveSpendJournalEntry(dbTx database.Tx, blockHash *chainhash.Hash) error {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
spendBucket := dbTx.Metadata().Bucket(spendJournalBucketName)
return spendBucket.Delete(blockHash[:])
}
// -----------------------------------------------------------------------------
// The unspent transaction output (utxo) set consists of an entry for each
// transaction which contains a utxo serialized using a format that is highly
// optimized to reduce space using domain specific compression algorithms. This
// format is a slightly modified version of the format used in Bitcoin Core.
//
// The serialized format is:
//
// <version><height><header code><unspentness bitmap>[<compressed txouts>,...]
//
// Field Type Size
// version VLQ variable
// block height VLQ variable
// header code VLQ variable
// unspentness bitmap []byte variable
// compressed txouts
// compressed amount VLQ variable
// compressed script []byte variable
//
// The serialized header code format is:
// bit 0 - containing transaction is a coinbase
// bit 1 - output zero is unspent
// bit 2 - output one is unspent
// bits 3-x - number of bytes in unspentness bitmap. When both bits 1 and 2
// are unset, it encodes N-1 since there must be at least one unspent
// output.
//
// The rationale for the header code scheme is as follows:
// - Transactions which only pay to a single output and a change output are
// extremely common, thus an extra byte for the unspentness bitmap can be
// avoided for them by encoding those two outputs in the low order bits.
// - Given it is encoded as a VLQ which can encode values up to 127 with a
// single byte, that leaves 4 bits to represent the number of bytes in the
// unspentness bitmap while still only consuming a single byte for the
// header code. In other words, an unspentness bitmap with up to 120
// transaction outputs can be encoded with a single-byte header code.
// This covers the vast majority of transactions.
// - Encoding N-1 bytes when both bits 1 and 2 are unset allows an additional
// 8 outpoints to be encoded before causing the header code to require an
// additional byte.
//
// Example 1:
// From tx in main blockchain:
// Blk 1, 0e3e2357e806b6cdb1f70b54c3a3a17b6714ee1f0e68bebb44a74b1efd512098
//
// 010103320496b538e853519c726a2c91e61ec11600ae1390813a627c66fb8be7947be63c52
// <><><><------------------------------------------------------------------>
// | | \--------\ |
// | height | compressed txout 0
// version header code
//
// - version: 1
// - height: 1
// - header code: 0x03 (coinbase, output zero unspent, 0 bytes of unspentness)
// - unspentness: Nothing since it is zero bytes
// - compressed txout 0:
// - 0x32: VLQ-encoded compressed amount for 5000000000 (50 BTC)
// - 0x04: special script type pay-to-pubkey
// - 0x96...52: x-coordinate of the pubkey
//
// Example 2:
// From tx in main blockchain:
// Blk 113931, 4a16969aa4764dd7507fc1de7f0baa4850a246de90c45e59a3207f9a26b5036f
//
// 0185f90b0a011200e2ccd6ec7c6e2e581349c77e067385fa8236bf8a800900b8025be1b3efc63b0ad48e7f9f10e87544528d58
// <><----><><><------------------------------------------><-------------------------------------------->
// | | | \-------------------\ | |
// version | \--------\ unspentness | compressed txout 2
// height header code compressed txout 0
//
// - version: 1
// - height: 113931
// - header code: 0x0a (output zero unspent, 1 byte in unspentness bitmap)
// - unspentness: [0x01] (bit 0 is set, so output 0+2 = 2 is unspent)
// NOTE: It's +2 since the first two outputs are encoded in the header code
// - compressed txout 0:
// - 0x12: VLQ-encoded compressed amount for 20000000 (0.2 BTC)
// - 0x00: special script type pay-to-pubkey-hash
// - 0xe2...8a: pubkey hash
// - compressed txout 2:
// - 0x8009: VLQ-encoded compressed amount for 15000000 (0.15 BTC)
// - 0x00: special script type pay-to-pubkey-hash
// - 0xb8...58: pubkey hash
//
// Example 3:
// From tx in main blockchain:
// Blk 338156, 1b02d1c8cfef60a189017b9a420c682cf4a0028175f2f563209e4ff61c8c3620
//
// 0193d06c100000108ba5b9e763011dd46a006572d820e448e12d2bbb38640bc718e6
// <><----><><----><-------------------------------------------------->
// | | | \-----------------\ |
// version | \--------\ unspentness |
// height header code compressed txout 22
//
// - version: 1
// - height: 338156
// - header code: 0x10 (2+1 = 3 bytes in unspentness bitmap)
// NOTE: It's +1 since neither bit 1 nor 2 are set, so N-1 is encoded.
// - unspentness: [0x00 0x00 0x10] (bit 20 is set, so output 20+2 = 22 is unspent)
// NOTE: It's +2 since the first two outputs are encoded in the header code
// - compressed txout 22:
// - 0x8ba5b9e763: VLQ-encoded compressed amount for 366875659 (3.66875659 BTC)
// - 0x01: special script type pay-to-script-hash
// - 0x1d...e6: script hash
// -----------------------------------------------------------------------------
// utxoEntryHeaderCode returns the calculated header code to be used when
// serializing the provided utxo entry and the number of bytes needed to encode
// the unspentness bitmap.
func utxoEntryHeaderCode(entry *UtxoEntry, highestOutputIndex uint32) (uint64, int, error) {
// The first two outputs are encoded separately, so offset the index
// accordingly to calculate the correct number of bytes needed to encode
// up to the highest unspent output index.
numBitmapBytes := int((highestOutputIndex + 6) / 8)
// As previously described, one less than the number of bytes is encoded
// when both output 0 and 1 are spent because there must be at least one
// unspent output. Adjust the number of bytes to encode accordingly and
// encode the value by shifting it over 3 bits.
output0Unspent := !entry.IsOutputSpent(0)
output1Unspent := !entry.IsOutputSpent(1)
var numBitmapBytesAdjustment int
if !output0Unspent && !output1Unspent {
if numBitmapBytes == 0 {
return 0, 0, AssertError("attempt to serialize utxo " +
"header for fully spent transaction")
}
numBitmapBytesAdjustment = 1
}
headerCode := uint64(numBitmapBytes-numBitmapBytesAdjustment) << 3
// Set the coinbase, output 0, and output 1 bits in the header code
// accordingly.
if entry.isCoinBase {
headerCode |= 0x01 // bit 0
}
if output0Unspent {
headerCode |= 0x02 // bit 1
}
if output1Unspent {
headerCode |= 0x04 // bit 2
}
return headerCode, numBitmapBytes, nil
}
// serializeUtxoEntry returns the entry serialized to a format that is suitable
// for long-term storage. The format is described in detail above.
func serializeUtxoEntry(entry *UtxoEntry) ([]byte, error) {
// Fully spent entries have no serialization.
if entry.IsFullySpent() {
return nil, nil
}
// Determine the output order by sorting the sparse output index keys.
outputOrder := make([]int, 0, len(entry.sparseOutputs))
for outputIndex := range entry.sparseOutputs {
outputOrder = append(outputOrder, int(outputIndex))
}
sort.Ints(outputOrder)
// Encode the header code and determine the number of bytes the
// unspentness bitmap needs.
highIndex := uint32(outputOrder[len(outputOrder)-1])
headerCode, numBitmapBytes, err := utxoEntryHeaderCode(entry, highIndex)
if err != nil {
return nil, err
}
// Calculate the size needed to serialize the entry.
size := serializeSizeVLQ(uint64(entry.version)) +
serializeSizeVLQ(uint64(entry.blockHeight)) +
serializeSizeVLQ(headerCode) + numBitmapBytes
for _, outputIndex := range outputOrder {
out := entry.sparseOutputs[uint32(outputIndex)]
if out.spent {
continue
}
size += compressedTxOutSize(uint64(out.amount), out.pkScript,
entry.version, out.compressed)
}
// Serialize the version, block height of the containing transaction,
// and header code.
serialized := make([]byte, size)
offset := putVLQ(serialized, uint64(entry.version))
offset += putVLQ(serialized[offset:], uint64(entry.blockHeight))
offset += putVLQ(serialized[offset:], headerCode)
// Serialize the unspentness bitmap.
for i := uint32(0); i < uint32(numBitmapBytes); i++ {
unspentBits := byte(0)
for j := uint32(0); j < 8; j++ {
// The first 2 outputs are encoded via the header code,
// so adjust the output index accordingly.
if !entry.IsOutputSpent(2 + i*8 + j) {
unspentBits |= 1 << uint8(j)
}
}
serialized[offset] = unspentBits
offset++
}
// Serialize the compressed unspent transaction outputs. Outputs that
// are already compressed are serialized without modifications.
for _, outputIndex := range outputOrder {
out := entry.sparseOutputs[uint32(outputIndex)]
if out.spent {
continue
}
offset += putCompressedTxOut(serialized[offset:],
uint64(out.amount), out.pkScript, entry.version,
out.compressed)
}
return serialized, nil
}
// deserializeUtxoEntry decodes a utxo entry from the passed serialized byte
// slice into a new UtxoEntry using a format that is suitable for long-term
// storage. The format is described in detail above.
func deserializeUtxoEntry(serialized []byte) (*UtxoEntry, error) {
// Deserialize the version.
version, bytesRead := deserializeVLQ(serialized)
offset := bytesRead
if offset >= len(serialized) {
return nil, errDeserialize("unexpected end of data after version")
}
// Deserialize the block height.
blockHeight, bytesRead := deserializeVLQ(serialized[offset:])
offset += bytesRead
if offset >= len(serialized) {
return nil, errDeserialize("unexpected end of data after height")
}
// Deserialize the header code.
code, bytesRead := deserializeVLQ(serialized[offset:])
offset += bytesRead
if offset >= len(serialized) {
return nil, errDeserialize("unexpected end of data after header")
}
// Decode the header code.
//
// Bit 0 indicates whether the containing transaction is a coinbase.
// Bit 1 indicates output 0 is unspent.
// Bit 2 indicates output 1 is unspent.
// Bits 3-x encodes the number of non-zero unspentness bitmap bytes that
// follow. When both output 0 and 1 are spent, it encodes N-1.
isCoinBase := code&0x01 != 0
output0Unspent := code&0x02 != 0
output1Unspent := code&0x04 != 0
numBitmapBytes := code >> 3
if !output0Unspent && !output1Unspent {
numBitmapBytes++
}
// Ensure there are enough bytes left to deserialize the unspentness
// bitmap.
if uint64(len(serialized[offset:])) < numBitmapBytes {
return nil, errDeserialize("unexpected end of data for " +
"unspentness bitmap")
}
// Create a new utxo entry with the details deserialized above to house
// all of the utxos.
entry := newUtxoEntry(int32(version), isCoinBase, int32(blockHeight))
// Add sparse output for unspent outputs 0 and 1 as needed based on the
// details provided by the header code.
var outputIndexes []uint32
if output0Unspent {
outputIndexes = append(outputIndexes, 0)
}
if output1Unspent {
outputIndexes = append(outputIndexes, 1)
}
// Decode the unspentness bitmap adding a sparse output for each unspent
// output.
for i := uint32(0); i < uint32(numBitmapBytes); i++ {
unspentBits := serialized[offset]
for j := uint32(0); j < 8; j++ {
if unspentBits&0x01 != 0 {
// The first 2 outputs are encoded via the
// header code, so adjust the output number
// accordingly.
outputNum := 2 + i*8 + j
outputIndexes = append(outputIndexes, outputNum)
}
unspentBits >>= 1
}
offset++
}
// Decode and add all of the utxos.
for i, outputIndex := range outputIndexes {
// Decode the next utxo. The script and amount fields of the
// utxo output are left compressed so decompression can be
// avoided on those that are not accessed. This is done since
// it is quite common for a redeeming transaction to only
// reference a single utxo from a referenced transaction.
compAmount, compScript, bytesRead, err := decodeCompressedTxOut(
serialized[offset:], int32(version))
if err != nil {
return nil, errDeserialize(fmt.Sprintf("unable to "+
"decode utxo at index %d: %v", i, err))
}
offset += bytesRead
entry.sparseOutputs[outputIndex] = &utxoOutput{
spent: false,
compressed: true,
pkScript: compScript,
amount: int64(compAmount),
}
}
return entry, nil
}
// dbFetchUtxoEntry uses an existing database transaction to fetch all unspent
// outputs for the provided Bitcoin transaction hash from the utxo set.
//
// When there is no entry for the provided hash, nil will be returned for the
// both the entry and the error.
func dbFetchUtxoEntry(dbTx database.Tx, hash *chainhash.Hash) (*UtxoEntry, error) {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Fetch the unspent transaction output information for the passed
// transaction hash. Return now when there is no entry.
utxoBucket := dbTx.Metadata().Bucket(utxoSetBucketName)
serializedUtxo := utxoBucket.Get(hash[:])
if serializedUtxo == nil {
return nil, nil
}
// A non-nil zero-length entry means there is an entry in the database
// for a fully spent transaction which should never be the case.
if len(serializedUtxo) == 0 {
return nil, AssertError(fmt.Sprintf("database contains entry "+
"for fully spent tx %v", hash))
}
// Deserialize the utxo entry and return it.
entry, err := deserializeUtxoEntry(serializedUtxo)
if err != nil {
// Ensure any deserialization errors are returned as database
// corruption errors.
if isDeserializeErr(err) {
return nil, database.Error{
ErrorCode: database.ErrCorruption,
Description: fmt.Sprintf("corrupt utxo entry "+
"for %v: %v", hash, err),
}
}
return nil, err
}
return entry, nil
}
// dbPutUtxoView uses an existing database transaction to update the utxo set
// in the database based on the provided utxo view contents and state. In
// particular, only the entries that have been marked as modified are written
// to the database.
func dbPutUtxoView(dbTx database.Tx, view *UtxoViewpoint) error {
utxoBucket := dbTx.Metadata().Bucket(utxoSetBucketName)
for txHashIter, entry := range view.entries {
// No need to update the database if the entry was not modified.
if entry == nil || !entry.modified {
continue
}
// Serialize the utxo entry without any entries that have been
// spent.
serialized, err := serializeUtxoEntry(entry)
if err != nil {
return err
}
// Make a copy of the hash because the iterator changes on each
// loop iteration and thus slicing it directly would cause the
// data to change out from under the put/delete funcs below.
txHash := txHashIter
// Remove the utxo entry if it is now fully spent.
if serialized == nil {
if err := utxoBucket.Delete(txHash[:]); err != nil {
return err
}
continue
}
// At this point the utxo entry is not fully spent, so store its
// serialization in the database.
err = utxoBucket.Put(txHash[:], serialized)
if err != nil {
return err
}
}
return nil
}
// -----------------------------------------------------------------------------
// The block index consists of two buckets with an entry for every block in the
// main chain. One bucket is for the hash to height mapping and the other is
// for the height to hash mapping.
//
// The serialized format for values in the hash to height bucket is:
// <height>
//
// Field Type Size
// height uint32 4 bytes
//
// The serialized format for values in the height to hash bucket is:
// <hash>
//
// Field Type Size
// hash chainhash.Hash chainhash.HashSize
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// -----------------------------------------------------------------------------
// dbPutBlockIndex uses an existing database transaction to update or add the
// block index entries for the hash to height and height to hash mappings for
// the provided values.
func dbPutBlockIndex(dbTx database.Tx, hash *chainhash.Hash, height int32) error {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Serialize the height for use in the index entries.
var serializedHeight [4]byte
byteOrder.PutUint32(serializedHeight[:], uint32(height))
// Add the block hash to height mapping to the index.
meta := dbTx.Metadata()
hashIndex := meta.Bucket(hashIndexBucketName)
if err := hashIndex.Put(hash[:], serializedHeight[:]); err != nil {
return err
}
// Add the block height to hash mapping to the index.
heightIndex := meta.Bucket(heightIndexBucketName)
return heightIndex.Put(serializedHeight[:], hash[:])
}
// dbRemoveBlockIndex uses an existing database transaction remove block index
// entries from the hash to height and height to hash mappings for the provided
// values.
func dbRemoveBlockIndex(dbTx database.Tx, hash *chainhash.Hash, height int32) error {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Remove the block hash to height mapping.
meta := dbTx.Metadata()
hashIndex := meta.Bucket(hashIndexBucketName)
if err := hashIndex.Delete(hash[:]); err != nil {
return err
}
// Remove the block height to hash mapping.
var serializedHeight [4]byte
byteOrder.PutUint32(serializedHeight[:], uint32(height))
heightIndex := meta.Bucket(heightIndexBucketName)
return heightIndex.Delete(serializedHeight[:])
}
// dbFetchHeightByHash uses an existing database transaction to retrieve the
// height for the provided hash from the index.
func dbFetchHeightByHash(dbTx database.Tx, hash *chainhash.Hash) (int32, error) {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
meta := dbTx.Metadata()
hashIndex := meta.Bucket(hashIndexBucketName)
serializedHeight := hashIndex.Get(hash[:])
if serializedHeight == nil {
str := fmt.Sprintf("block %s is not in the main chain", hash)
return 0, errNotInMainChain(str)
}
return int32(byteOrder.Uint32(serializedHeight)), nil
}
// dbFetchHashByHeight uses an existing database transaction to retrieve the
// hash for the provided height from the index.
func dbFetchHashByHeight(dbTx database.Tx, height int32) (*chainhash.Hash, error) {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
var serializedHeight [4]byte
byteOrder.PutUint32(serializedHeight[:], uint32(height))
meta := dbTx.Metadata()
heightIndex := meta.Bucket(heightIndexBucketName)
hashBytes := heightIndex.Get(serializedHeight[:])
if hashBytes == nil {
str := fmt.Sprintf("no block at height %d exists", height)
return nil, errNotInMainChain(str)
}
var hash chainhash.Hash
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
copy(hash[:], hashBytes)
return &hash, nil
}
// -----------------------------------------------------------------------------
// The best chain state consists of the best block hash and height, the total
// number of transactions up to and including those in the best block, and the
// accumulated work sum up to and including the best block.
//
// The serialized format is:
//
// <block hash><block height><total txns><work sum length><work sum>
//
// Field Type Size
// block hash chainhash.Hash chainhash.HashSize
// block height uint32 4 bytes
// total txns uint64 8 bytes
// work sum length uint32 4 bytes
// work sum big.Int work sum length
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// -----------------------------------------------------------------------------
// bestChainState represents the data to be stored the database for the current
// best chain state.
type bestChainState struct {
hash chainhash.Hash
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
height uint32
totalTxns uint64
workSum *big.Int
}
// serializeBestChainState returns the serialization of the passed block best
// chain state. This is data to be stored in the chain state bucket.
func serializeBestChainState(state bestChainState) []byte {
// Calculate the full size needed to serialize the chain state.
workSumBytes := state.workSum.Bytes()
workSumBytesLen := uint32(len(workSumBytes))
serializedLen := chainhash.HashSize + 4 + 8 + 4 + workSumBytesLen
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Serialize the chain state.
serializedData := make([]byte, serializedLen)
copy(serializedData[0:chainhash.HashSize], state.hash[:])
offset := uint32(chainhash.HashSize)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
byteOrder.PutUint32(serializedData[offset:], state.height)
offset += 4
byteOrder.PutUint64(serializedData[offset:], state.totalTxns)
offset += 8
byteOrder.PutUint32(serializedData[offset:], workSumBytesLen)
offset += 4
copy(serializedData[offset:], workSumBytes)
return serializedData[:]
}
// deserializeBestChainState deserializes the passed serialized best chain
// state. This is data stored in the chain state bucket and is updated after
// every block is connected or disconnected form the main chain.
// block.
func deserializeBestChainState(serializedData []byte) (bestChainState, error) {
// Ensure the serialized data has enough bytes to properly deserialize
// the hash, height, total transactions, and work sum length.
if len(serializedData) < chainhash.HashSize+16 {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return bestChainState{}, database.Error{
ErrorCode: database.ErrCorruption,
Description: "corrupt best chain state",
}
}
state := bestChainState{}
copy(state.hash[:], serializedData[0:chainhash.HashSize])
offset := uint32(chainhash.HashSize)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
state.height = byteOrder.Uint32(serializedData[offset : offset+4])
offset += 4
state.totalTxns = byteOrder.Uint64(serializedData[offset : offset+8])
offset += 8
workSumBytesLen := byteOrder.Uint32(serializedData[offset : offset+4])
offset += 4
// Ensure the serialized data has enough bytes to deserialize the work
// sum.
if uint32(len(serializedData[offset:])) < workSumBytesLen {
return bestChainState{}, database.Error{
ErrorCode: database.ErrCorruption,
Description: "corrupt best chain state",
}
}
workSumBytes := serializedData[offset : offset+workSumBytesLen]
state.workSum = new(big.Int).SetBytes(workSumBytes)
return state, nil
}
// dbPutBestState uses an existing database transaction to update the best chain
// state with the given parameters.
func dbPutBestState(dbTx database.Tx, snapshot *BestState, workSum *big.Int) error {
// Serialize the current best chain state.
serializedData := serializeBestChainState(bestChainState{
hash: snapshot.Hash,
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
height: uint32(snapshot.Height),
totalTxns: snapshot.TotalTxns,
workSum: workSum,
})
// Store the current best chain state into the database.
return dbTx.Metadata().Put(chainStateKeyName, serializedData)
}
// createChainState initializes both the database and the chain state to the
// genesis block. This includes creating the necessary buckets and inserting
// the genesis block, so it must only be called on an uninitialized database.
func (b *BlockChain) createChainState() error {
// Create a new node from the genesis block and set it as the best node.
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
genesisBlock := btcutil.NewBlock(b.chainParams.GenesisBlock)
header := &genesisBlock.MsgBlock().Header
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
node := newBlockNode(header, 0)
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
b.bestChain.SetTip(node)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Add the new node to the index which is used for faster lookups.
b.index.AddNode(node)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Initialize the state related to the best block. Since it is the
// genesis block, use its timestamp for the median time.
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
numTxns := uint64(len(genesisBlock.MsgBlock().Transactions))
blockSize := uint64(genesisBlock.MsgBlock().SerializeSize())
blockWeight := uint64(GetBlockWeight(genesisBlock))
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
b.stateSnapshot = newBestState(node, blockSize, blockWeight, numTxns,
numTxns, time.Unix(node.timestamp, 0))
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Create the initial the database chain state including creating the
// necessary index buckets and inserting the genesis block.
err := b.db.Update(func(dbTx database.Tx) error {
// Create the bucket that houses the chain block hash to height
// index.
meta := dbTx.Metadata()
_, err := meta.CreateBucket(hashIndexBucketName)
if err != nil {
return err
}
// Create the bucket that houses the chain block height to hash
// index.
_, err = meta.CreateBucket(heightIndexBucketName)
if err != nil {
return err
}
// Create the bucket that houses the spend journal data.
_, err = meta.CreateBucket(spendJournalBucketName)
if err != nil {
return err
}
// Create the bucket that houses the utxo set. Note that the
// genesis block coinbase transaction is intentionally not
// inserted here since it is not spendable by consensus rules.
_, err = meta.CreateBucket(utxoSetBucketName)
if err != nil {
return err
}
// Add the genesis block hash to height and height to hash
// mappings to the index.
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
err = dbPutBlockIndex(dbTx, &node.hash, node.height)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
if err != nil {
return err
}
// Store the current best chain state into the database.
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
err = dbPutBestState(dbTx, b.stateSnapshot, node.workSum)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
if err != nil {
return err
}
// Store the genesis block into the database.
return dbTx.StoreBlock(genesisBlock)
})
return err
}
// initChainState attempts to load and initialize the chain state from the
// database. When the db does not yet contain any chain state, both it and the
// chain state are initialized to the genesis block.
func (b *BlockChain) initChainState() error {
// Attempt to load the chain state from the database.
var isStateInitialized bool
err := b.db.View(func(dbTx database.Tx) error {
// Fetch the stored chain state from the database metadata.
// When it doesn't exist, it means the database hasn't been
// initialized for use with chain yet, so break out now to allow
// that to happen under a writable database transaction.
serializedData := dbTx.Metadata().Get(chainStateKeyName)
if serializedData == nil {
return nil
}
log.Tracef("Serialized chain state: %x", serializedData)
state, err := deserializeBestChainState(serializedData)
if err != nil {
return err
}
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
// Load all of the headers from the data for the known best
// chain and construct the block index accordingly. Since the
// number of nodes are already known, perform a single alloc
// for them versus a whole bunch of little ones to reduce
// pressure on the GC.
log.Infof("Loading block index. This might take a while...")
bestHeight := int32(state.height)
blockNodes := make([]blockNode, bestHeight+1)
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
var tip *blockNode
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
for height := int32(0); height <= bestHeight; height++ {
header, err := dbFetchHeaderByHeight(dbTx, height)
if err != nil {
return err
}
// Initialize the block node for the block, connect it,
// and add it to the block index.
node := &blockNodes[height]
initBlockNode(node, header, height)
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
if tip != nil {
node.parent = tip
node.workSum = node.workSum.Add(tip.workSum,
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
node.workSum)
}
b.index.AddNode(node)
// This node is now the end of the best chain.
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
tip = node
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
}
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
// Ensure the resulting best chain matches the stored best state
// hash and set the best chain view accordingly.
if tip == nil || tip.hash != state.hash {
var tipHash chainhash.Hash
if tip != nil {
tipHash = tip.hash
}
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
return AssertError(fmt.Sprintf("initChainState: block "+
"index chain tip %s does not match stored "+
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
"best state %s", tipHash, state.hash))
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
}
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
b.bestChain.SetTip(tip)
blockchain: Convert to full block index in mem. This reworks the block index code such that it loads all of the headers in the main chain at startup and constructs the full block index accordingly. Since the full index from the current best tip all the way back to the genesis block is now guaranteed to be in memory, this also removes all code related to dynamically loading the nodes and updates some of the logic to take advantage of the fact traversing the block index can longer potentially fail. There are also more optimizations and simplifications that can be made in the future as a result of this. Due to removing all of the extra overhead of tracking the dynamic state, and ensuring the block node structs are aligned to eliminate extra padding, the end result of a fully populated block index now takes quite a bit less memory than the previous dynamically loaded version. The main downside is that it now takes a while to start whereas it was nearly instant before, however, it is much better to provide more efficient runtime operation since that is its ultimate purpose and the benefits far outweigh this downside. Some benefits are: - Since every block node is in memory, the recent code which reconstructs headers from block nodes means that all headers can always be served from memory which is important since the majority of the network has moved to header-based semantics - Several of the error paths can be removed since they are no longer necessary - It is no longer expensive to calculate CSV sequence locks or median times of blocks way in the past - It will be possible to create much more efficient iteration and simplified views of the overall index - The entire threshold state database cache can be removed since it is cheap to construct it from the full block index as needed An overview of the logic changes are as follows: - Move AncestorNode from blockIndex to blockNode and greatly simplify since it no longer has to deal with the possibility of dynamically loading nodes and related failures - Rename RelativeNode to RelativeAncestor, move to blockNode, and redefine in terms of AncestorNode - Move CalcPastMedianTime from blockIndex to blockNode and remove no longer necessary test for nil - Change calcSequenceLock to use Ancestor instead of RelativeAncestor since it reads more clearly
2017-02-03 19:13:53 +01:00
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Load the raw block bytes for the best block.
blockBytes, err := dbTx.FetchBlock(&state.hash)
if err != nil {
return err
}
var block wire.MsgBlock
err = block.Deserialize(bytes.NewReader(blockBytes))
if err != nil {
return err
}
// Initialize the state related to the best block.
blockSize := uint64(len(blockBytes))
blockWeight := uint64(GetBlockWeight(btcutil.NewBlock(&block)))
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
numTxns := uint64(len(block.Transactions))
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
b.stateSnapshot = newBestState(tip, blockSize, blockWeight,
numTxns, state.totalTxns, tip.CalcPastMedianTime())
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
isStateInitialized = true
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return nil
})
if err != nil {
return err
}
// There is nothing more to do if the chain state was initialized.
if isStateInitialized {
return nil
}
// At this point the database has not already been initialized, so
// initialize both it and the chain state to the genesis block.
return b.createChainState()
}
// dbFetchHeaderByHash uses an existing database transaction to retrieve the
// block header for the provided hash.
func dbFetchHeaderByHash(dbTx database.Tx, hash *chainhash.Hash) (*wire.BlockHeader, error) {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
headerBytes, err := dbTx.FetchBlockHeader(hash)
if err != nil {
return nil, err
}
var header wire.BlockHeader
err = header.Deserialize(bytes.NewReader(headerBytes))
if err != nil {
return nil, err
}
return &header, nil
}
// dbFetchHeaderByHeight uses an existing database transaction to retrieve the
// block header for the provided height.
func dbFetchHeaderByHeight(dbTx database.Tx, height int32) (*wire.BlockHeader, error) {
hash, err := dbFetchHashByHeight(dbTx, height)
if err != nil {
return nil, err
}
return dbFetchHeaderByHash(dbTx, hash)
}
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
// dbFetchBlockByNode uses an existing database transaction to retrieve the
// raw block for the provided node, deserialize it, and return a btcutil.Block
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// with the height set.
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
func dbFetchBlockByNode(dbTx database.Tx, node *blockNode) (*btcutil.Block, error) {
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
// Load the raw block bytes from the database.
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
blockBytes, err := dbTx.FetchBlock(&node.hash)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
if err != nil {
return nil, err
}
// Create the encapsulated block and set the height appropriately.
block, err := btcutil.NewBlockFromBytes(blockBytes)
if err != nil {
return nil, err
}
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
block.SetHeight(node.height)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return block, nil
}
// BlockByHeight returns the block at the given height in the main chain.
//
// This function is safe for concurrent access.
func (b *BlockChain) BlockByHeight(blockHeight int32) (*btcutil.Block, error) {
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
// Lookup the block height in the best chain.
node := b.bestChain.NodeByHeight(blockHeight)
if node == nil {
str := fmt.Sprintf("no block at height %d exists", blockHeight)
return nil, errNotInMainChain(str)
}
// Load the block from the database and return it.
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
var block *btcutil.Block
err := b.db.View(func(dbTx database.Tx) error {
var err error
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
block, err = dbFetchBlockByNode(dbTx, node)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return err
})
return block, err
}
// BlockByHash returns the block from the main chain with the given hash with
// the appropriate chain height set.
//
// This function is safe for concurrent access.
func (b *BlockChain) BlockByHash(hash *chainhash.Hash) (*btcutil.Block, error) {
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
// Lookup the block hash in block index and ensure it is in the best
// chain.
node := b.index.LookupNode(hash)
if node == nil || !b.bestChain.Contains(node) {
str := fmt.Sprintf("block %s is not in the main chain", hash)
return nil, errNotInMainChain(str)
}
// Load the block from the database and return it.
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
var block *btcutil.Block
err := b.db.View(func(dbTx database.Tx) error {
var err error
blockchain: Refactor to use new chain view. - Remove inMainChain from block nodes since that can now be efficiently determined by using the chain view - Track the best chain via a chain view instead of a single block node - Use the tip of the best chain view everywhere bestNode was used - Update chain view tip instead of updating best node - Change reorg logic to use more efficient chain view fork finding logic - Change block locator code over to use more efficient chain view logic - Remove now unused block-index-based block locator code - Move BlockLocator definition to chain.go - Move BlockLocatorFromHash and LatestBlockLocator to chain.go - Update both to use more efficient chain view logic - Rework IsCheckpointCandidate to use block index and chain view - Optimize MainChainHasBlock to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHeightByHash to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize BlockHashByHeight to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Removed error return since it can no longer fail - Optimize HeightRange to use chain view instead of hitting db - Move to chain.go since it no longer involves database I/O - Optimize BlockByHeight to use chain view for main chain check - Optimize BlockByHash to use chain view for main chain check
2017-08-18 14:25:54 +02:00
block, err = dbFetchBlockByNode(dbTx, node)
blockchain: Rework to use new db interface. This commit is the first stage of several that are planned to convert the blockchain package into a concurrent safe package that will ultimately allow support for multi-peer download and concurrent chain processing. The goal is to update btcd proper after each step so it can take advantage of the enhancements as they are developed. In addition to the aforementioned benefit, this staged approach has been chosen since it is absolutely critical to maintain consensus. Separating the changes into several stages makes it easier for reviewers to logically follow what is happening and therefore helps prevent consensus bugs. Naturally there are significant automated tests to help prevent consensus issues as well. The main focus of this stage is to convert the blockchain package to use the new database interface and implement the chain-related functionality which it no longer handles. It also aims to improve efficiency in various areas by making use of the new database and chain capabilities. The following is an overview of the chain changes: - Update to use the new database interface - Add chain-related functionality that the old database used to handle - Main chain structure and state - Transaction spend tracking - Implement a new pruned unspent transaction output (utxo) set - Provides efficient direct access to the unspent transaction outputs - Uses a domain specific compression algorithm that understands the standard transaction scripts in order to significantly compress them - Removes reliance on the transaction index and paves the way toward eventually enabling block pruning - Modify the New function to accept a Config struct instead of inidividual parameters - Replace the old TxStore type with a new UtxoViewpoint type that makes use of the new pruned utxo set - Convert code to treat the new UtxoViewpoint as a rolling view that is used between connects and disconnects to improve efficiency - Make best chain state always set when the chain instance is created - Remove now unnecessary logic for dealing with unset best state - Make all exported functions concurrent safe - Currently using a single chain state lock as it provides a straight forward and easy to review path forward however this can be improved with more fine grained locking - Optimize various cases where full blocks were being loaded when only the header is needed to help reduce the I/O load - Add the ability for callers to get a snapshot of the current best chain stats in a concurrent safe fashion - Does not block callers while new blocks are being processed - Make error messages that reference transaction outputs consistently use <transaction hash>:<output index> - Introduce a new AssertError type an convert internal consistency checks to use it - Update tests and examples to reflect the changes - Add a full suite of tests to ensure correct functionality of the new code The following is an overview of the btcd changes: - Update to use the new database and chain interfaces - Temporarily remove all code related to the transaction index - Temporarily remove all code related to the address index - Convert all code that uses transaction stores to use the new utxo view - Rework several calls that required the block manager for safe concurrency to use the chain package directly now that it is concurrent safe - Change all calls to obtain the best hash to use the new best state snapshot capability from the chain package - Remove workaround for limits on fetching height ranges since the new database interface no longer imposes them - Correct the gettxout RPC handler to return the best chain hash as opposed the hash the txout was found in - Optimize various RPC handlers: - Change several of the RPC handlers to use the new chain snapshot capability to avoid needlessly loading data - Update several handlers to use new functionality to avoid accessing the block manager so they are able to return the data without blocking when the server is busy processing blocks - Update non-verbose getblock to avoid deserialization and serialization overhead - Update getblockheader to request the block height directly from chain and only load the header - Update getdifficulty to use the new cached data from chain - Update getmininginfo to use the new cached data from chain - Update non-verbose getrawtransaction to avoid deserialization and serialization overhead - Update gettxout to use the new utxo store versus loading full transactions using the transaction index The following is an overview of the utility changes: - Update addblock to use the new database and chain interfaces - Update findcheckpoint to use the new database and chain interfaces - Remove the dropafter utility which is no longer supported NOTE: The transaction index and address index will be reimplemented in another commit.
2015-08-26 06:03:18 +02:00
return err
})
return block, err
}