What causes "Could not truncate current file to known size in db: %s" in hyperledger/fabric?

Restore the block files to match the index DB checkpoint (same backup) so the recorded latestFileSize is valid

Could not truncate current file to known size in db: %s

Error message

Could not truncate current file to known size in db: %s

What it means

Panic when currentFileWriter.truncateFile cannot truncate the current block file to latestFileSize recorded in the DB checkpoint, ensuring the file length matches committed data. A mismatch means uncommitted bytes from a crash would be appended to, so startup aborts.

Source

Thrown at common/ledger/blkstorage/blockfile_mgr.go:129

			panic(fmt.Sprintf("Could not build blockfilesInfo info from block files: %s", err))
		}
		logger.Debugf("Info constructed by scanning the blocks dir = %s", spew.Sdump(blockfilesInfo))
	} else {
		logger.Debug(`Syncing block information from block storage (if needed)`)
		syncBlockfilesInfoFromFS(rootDir, blockfilesInfo)
	}
	err = mgr.saveBlkfilesInfo(blockfilesInfo, true)
	if err != nil {
		panic(fmt.Sprintf("Could not save next block file info to db: %s", err))
	}

	currentFileWriter, err := newBlockfileWriter(deriveBlockfilePath(rootDir, blockfilesInfo.latestFileNumber))
	if err != nil {
		panic(fmt.Sprintf("Could not open writer to current file: %s", err))
	}
	err = currentFileWriter.truncateFile(blockfilesInfo.latestFileSize)
	if err != nil {
		panic(fmt.Sprintf("Could not truncate current file to known size in db: %s", err))
	}
	if mgr.index, err = newBlockIndex(indexConfig, indexStore); err != nil {
		panic(fmt.Sprintf("error in block index: %s", err))
	}

	mgr.blockfilesInfo = blockfilesInfo
	bsi, err := loadBootstrappingSnapshotInfo(rootDir)
	if err != nil {
		return nil, err
	}
	mgr.bootstrappingSnapshotInfo = bsi
	mgr.currentFileWriter = currentFileWriter
	mgr.blkfilesInfoCond = sync.NewCond(&sync.Mutex{})

	if err := mgr.syncIndex(); err != nil {
		return nil, err
	}

View on GitHub (pinned to 2736b63f8f)

Solutions

Restore the block files to match the index DB checkpoint (same backup) so the recorded latestFileSize is valid
If the FS is the source of truth, reset/sync the index DB (delete index so the peer rescans with constructBlockfilesInfo)
Fix permissions and filesystem health, then restart the peer

Example fix

// before: restoring chains files from an older backup than the index DB
// after: restore both dirs atomically from one snapshot
// systemctl stop peer && rsync -a snapshot/ledgersData/ ledgersData/ && systemctl start peer

Defensive patterns

Strategy: validation

Validate before calling

fi, err := os.Stat(deriveBlockfilePath(rootDir, latestFileNumber))
if err != nil { return err }
if fi.Size() < recordedLatestFileSize {
    return fmt.Errorf("block file %d smaller (%d) than db checkpoint (%d)", latestFileNumber, fi.Size(), recordedLatestFileSize)
}

Try / catch

defer func() {
    if r := recover(); r != nil {
        err = fmt.Errorf("truncate to checkpoint size failed: %v", r)
    }
}()

Prevention

Restore index DB and block files from the same backup point
Take backups only when the peer is stopped
Keep block files writable by the peer user
Monitor storage for I/O errors

When it happens

Trigger: truncateFile fails (I/O error, permission denied, file shorter/absent, read-only fs) aligning the block file to blockfilesInfo.latestFileSize in newBlockfileMgr.

Common situations: Block file missing or shrunk after a partial restore/backup mismatch (DB says file is 50MB, FS has 10MB); disk errors; permission issues on the chains dir.

Related errors

AI-assisted analysis of hyperledger/fabric@2736b63f8f (2026-09-04). Data as JSON: /api/errors/884e1093a570fbb1. Report an issue: GitHub.