lbcd/util/addblock/import.go
Dave Collins 6222b1d8cc Rework and Improve addblock utility.
The addblock utility was originally written as a quick debug tool during
initial development to populate blocks into the database.  However, now
that it has been designated as the standard way to import bootstrap.dat
(and indeed block data files in general), it was lacking a few features
such as properly checking against the chain rules and known good
checkpoints.

This commit reworks and improves the utility in several ways:

- Imported blocks are now checked against the chain rules including
  checkpoints to ensure they match the known good chain
- The utility now properly shuts down after processing all blocks
- Attempting to import orphan blocks (blocks which build off a block you
  don't yet have in the database) returns an error
- Blocks that are already known are now skipped instead of causing an
  error which means you can stop and restart the import mid-way through
  without issues or start it after you've already downloaded a
  portion of the chain
- The block height is no longer assumed to start at 0 which means input
  files that start later in the chain will work properly so long as you
  already have the chain at least up to the point of the block just before
  the first one in the input file
- Improved error handling and reporting
- How often the progress display is shown is now configurable
- Statistics about how many blocks were processed, imported, and already
  known are now displayed after the input file has been fully processed

This resolves comments made in #60.
2014-01-14 01:18:52 -06:00

252 lines
7 KiB
Go

// Copyright (c) 2013-2014 Conformal Systems LLC.
// Use of this source code is governed by an ISC
// license that can be found in the LICENSE file.
package main
import (
"encoding/binary"
"fmt"
"github.com/conformal/btcchain"
"github.com/conformal/btcdb"
_ "github.com/conformal/btcdb/ldb"
"github.com/conformal/btcutil"
"github.com/conformal/btcwire"
"io"
"sync"
)
var zeroHash = btcwire.ShaHash{}
// importResults houses the stats and result as an import operation.
type importResults struct {
blocksProcessed int64
blocksImported int64
err error
}
// blockImporter houses information about an ongoing import from a block data
// file to the block database.
type blockImporter struct {
db btcdb.Db
chain *btcchain.BlockChain
r io.ReadSeeker
processQueue chan []byte
doneChan chan bool
errChan chan error
quit chan bool
wg sync.WaitGroup
blocksProcessed int64
blocksImported int64
}
// readBlock reads the next block from the input file.
func (bi *blockImporter) readBlock() ([]byte, error) {
// The block file format is:
// <network> <block length> <serialized block>
var net uint32
err := binary.Read(bi.r, binary.LittleEndian, &net)
if err != nil {
if err != io.EOF {
return nil, err
}
// No block and no error means there are no more blocks to read.
return nil, nil
}
if net != uint32(activeNetwork) {
return nil, fmt.Errorf("network mismatch -- got %x, want %x",
net, uint32(activeNetwork))
}
// Read the block length and ensure it is sane.
var blockLen uint32
if err := binary.Read(bi.r, binary.LittleEndian, &blockLen); err != nil {
return nil, err
}
if blockLen > btcwire.MaxBlockPayload {
return nil, fmt.Errorf("block payload of %d bytes is larger "+
"than the max allowed %d bytes", blockLen,
btcwire.MaxBlockPayload)
}
serializedBlock := make([]byte, blockLen)
if _, err := io.ReadFull(bi.r, serializedBlock); err != nil {
return nil, err
}
return serializedBlock, nil
}
// processBlock potentially imports the block into the database. It first
// deserializes the raw block while checking for errors. Already known blocks
// are skipped and orphan blocks are considered errors. Finally, it runs the
// block through the chain rules to ensure it follows all rules and matches
// up to the known checkpoint. Returns whether the block was imported along
// with any potential errors.
func (bi *blockImporter) processBlock(serializedBlock []byte) (bool, error) {
// Deserialize the block which includes checks for malformed blocks.
block, err := btcutil.NewBlockFromBytes(serializedBlock)
if err != nil {
return false, err
}
blockSha, err := block.Sha()
if err != nil {
return false, err
}
// Skip blocks that already exist.
if bi.db.ExistsSha(blockSha) {
return false, nil
}
// Don't bother trying to process orphans.
prevHash := &block.MsgBlock().Header.PrevBlock
if !prevHash.IsEqual(&zeroHash) && !bi.db.ExistsSha(prevHash) {
return false, fmt.Errorf("import file contains block %v which "+
"does not link to the available block chain", blockSha)
}
// Ensure the blocks follows all of the chain rules and match up to the
// known checkpoints.
if err := bi.chain.ProcessBlock(block, true); err != nil {
return false, err
}
return true, nil
}
// readHandler is the main handler for reading blocks from the import file.
// This allows block processing to take place in parallel with block reads.
// It must be run as a goroutine.
func (bi *blockImporter) readHandler() {
out:
for {
// Read the next block from the file and if anything goes wrong
// notify the status handler with the error and bail.
serializedBlock, err := bi.readBlock()
if err != nil {
bi.errChan <- fmt.Errorf("Error reading from input "+
"file: %v", err.Error())
break out
}
// A nil block with no error means we're done.
if serializedBlock == nil {
break out
}
// Send the block or quit if we've been signalled to exit by
// the status handler due to an error elsewhere.
select {
case bi.processQueue <- serializedBlock:
case <-bi.quit:
break out
}
}
// Close the processing channel to signal no more blocks are coming.
close(bi.processQueue)
bi.wg.Done()
}
// processHandler is the main handler for processing blocks. This allows block
// processing to take place in parallel with block reads from the import file.
// It must be run as a goroutine.
func (bi *blockImporter) processHandler() {
out:
for {
select {
case serializedBlock, ok := <-bi.processQueue:
// We're done when the channel is closed.
if !ok {
break out
}
bi.blocksProcessed++
imported, err := bi.processBlock(serializedBlock)
if err != nil {
bi.errChan <- err
break out
}
if imported {
bi.blocksImported++
}
if cfg.Progress != 0 && bi.blocksProcessed > 0 &&
bi.blocksProcessed%int64(cfg.Progress) == 0 {
log.Infof("Processed %d blocks", bi.blocksProcessed)
}
case <-bi.quit:
break out
}
}
bi.wg.Done()
}
// statusHandler waits for updates from the import operation and notifies
// the passed doneChan with the results of the import. It also causes all
// goroutines to exit if an error is reported from any of them.
func (bi *blockImporter) statusHandler(resultsChan chan *importResults) {
select {
// An error from either of the goroutines means we're done so signal
// caller with the error and signal all goroutines to quit.
case err := <-bi.errChan:
resultsChan <- &importResults{
blocksProcessed: bi.blocksProcessed,
blocksImported: bi.blocksImported,
err: err,
}
close(bi.quit)
// The import finished normally.
case <-bi.doneChan:
resultsChan <- &importResults{
blocksProcessed: bi.blocksProcessed,
blocksImported: bi.blocksImported,
err: nil,
}
}
}
// Import is the core function which handles importing the blocks from the file
// associated with the block importer to the database. It returns a channel
// on which the results will be returned when the operation has completed.
func (bi *blockImporter) Import() chan *importResults {
// Start up the read and process handling goroutines. This setup allows
// blocks to be read from disk in parallel while being processed.
bi.wg.Add(2)
go bi.readHandler()
go bi.processHandler()
// Wait for the import to finish in a separate goroutine and signal
// the status handler when done.
go func() {
bi.wg.Wait()
bi.doneChan <- true
}()
// Start the status handler and return the result the channel that it
// will send the results on when the import is done.
resultChan := make(chan *importResults)
go bi.statusHandler(resultChan)
return resultChan
}
// newBlockImporter returns a new importer for the provided file reader seeker
// and database.
func newBlockImporter(db btcdb.Db, r io.ReadSeeker) *blockImporter {
return &blockImporter{
db: db,
r: r,
processQueue: make(chan []byte, 2),
doneChan: make(chan bool),
errChan: make(chan error),
quit: make(chan bool),
chain: btcchain.New(db, activeNetwork, nil),
}
}