restic/cmd/restic/cmd_copy.go
MichaelEischer 2c3e5d943d
Merge pull request #3593 from DarkKirb/parallelize-restic-copy
Parallelize blob upload/download for restic copy
2021-12-29 22:31:54 +01:00

257 lines
7.5 KiB
Go

package main
import (
"context"
"fmt"
"github.com/restic/restic/internal/debug"
"github.com/restic/restic/internal/restic"
"golang.org/x/sync/errgroup"
"github.com/spf13/cobra"
)
var cmdCopy = &cobra.Command{
Use: "copy [flags] [snapshotID ...]",
Short: "Copy snapshots from one repository to another",
Long: `
The "copy" command copies one or more snapshots from one repository to another.
NOTE: This process will have to both download (read) and upload (write) the
entire snapshot(s) due to the different encryption keys used in the source and
destination repositories. This /may incur higher bandwidth usage and costs/ than
expected during normal backup runs.
NOTE: The copying process does not re-chunk files, which may break deduplication
between the files copied and files already stored in the destination repository.
This means that copied files, which existed in both the source and destination
repository, /may occupy up to twice their space/ in the destination repository.
This can be mitigated by the "--copy-chunker-params" option when initializing a
new destination repository using the "init" command.
`,
RunE: func(cmd *cobra.Command, args []string) error {
return runCopy(copyOptions, globalOptions, args)
},
}
// CopyOptions bundles all options for the copy command.
type CopyOptions struct {
secondaryRepoOptions
Hosts []string
Tags restic.TagLists
Paths []string
}
var copyOptions CopyOptions
func init() {
cmdRoot.AddCommand(cmdCopy)
f := cmdCopy.Flags()
initSecondaryRepoOptions(f, &copyOptions.secondaryRepoOptions, "destination", "to copy snapshots to")
f.StringArrayVarP(&copyOptions.Hosts, "host", "H", nil, "only consider snapshots for this `host`, when no snapshot ID is given (can be specified multiple times)")
f.Var(&copyOptions.Tags, "tag", "only consider snapshots which include this `taglist`, when no snapshot ID is given")
f.StringArrayVar(&copyOptions.Paths, "path", nil, "only consider snapshots which include this (absolute) `path`, when no snapshot ID is given")
}
func runCopy(opts CopyOptions, gopts GlobalOptions, args []string) error {
dstGopts, err := fillSecondaryGlobalOpts(opts.secondaryRepoOptions, gopts, "destination")
if err != nil {
return err
}
ctx, cancel := context.WithCancel(gopts.ctx)
defer cancel()
srcRepo, err := OpenRepository(gopts)
if err != nil {
return err
}
dstRepo, err := OpenRepository(dstGopts)
if err != nil {
return err
}
if !gopts.NoLock {
srcLock, err := lockRepo(ctx, srcRepo)
defer unlockRepo(srcLock)
if err != nil {
return err
}
}
dstLock, err := lockRepo(ctx, dstRepo)
defer unlockRepo(dstLock)
if err != nil {
return err
}
debug.Log("Loading source index")
if err := srcRepo.LoadIndex(ctx); err != nil {
return err
}
debug.Log("Loading destination index")
if err := dstRepo.LoadIndex(ctx); err != nil {
return err
}
dstSnapshotByOriginal := make(map[restic.ID][]*restic.Snapshot)
for sn := range FindFilteredSnapshots(ctx, dstRepo, opts.Hosts, opts.Tags, opts.Paths, nil) {
if sn.Original != nil && !sn.Original.IsNull() {
dstSnapshotByOriginal[*sn.Original] = append(dstSnapshotByOriginal[*sn.Original], sn)
}
// also consider identical snapshot copies
dstSnapshotByOriginal[*sn.ID()] = append(dstSnapshotByOriginal[*sn.ID()], sn)
}
// remember already processed trees across all snapshots
visitedTrees := restic.NewIDSet()
for sn := range FindFilteredSnapshots(ctx, srcRepo, opts.Hosts, opts.Tags, opts.Paths, args) {
Verbosef("\nsnapshot %s of %v at %s)\n", sn.ID().Str(), sn.Paths, sn.Time)
// check whether the destination has a snapshot with the same persistent ID which has similar snapshot fields
srcOriginal := *sn.ID()
if sn.Original != nil {
srcOriginal = *sn.Original
}
if originalSns, ok := dstSnapshotByOriginal[srcOriginal]; ok {
isCopy := false
for _, originalSn := range originalSns {
if similarSnapshots(originalSn, sn) {
Verbosef("skipping source snapshot %s, was already copied to snapshot %s\n", sn.ID().Str(), originalSn.ID().Str())
isCopy = true
break
}
}
if isCopy {
continue
}
}
Verbosef(" copy started, this may take a while...\n")
if err := copyTree(ctx, srcRepo, dstRepo, visitedTrees, *sn.Tree); err != nil {
return err
}
debug.Log("tree copied")
if err = dstRepo.Flush(ctx); err != nil {
return err
}
debug.Log("flushed packs and saved index")
// save snapshot
sn.Parent = nil // Parent does not have relevance in the new repo.
// Use Original as a persistent snapshot ID
if sn.Original == nil {
sn.Original = sn.ID()
}
newID, err := dstRepo.SaveJSONUnpacked(ctx, restic.SnapshotFile, sn)
if err != nil {
return err
}
Verbosef("snapshot %s saved\n", newID.Str())
}
return nil
}
func similarSnapshots(sna *restic.Snapshot, snb *restic.Snapshot) bool {
// everything except Parent and Original must match
if !sna.Time.Equal(snb.Time) || !sna.Tree.Equal(*snb.Tree) || sna.Hostname != snb.Hostname ||
sna.Username != snb.Username || sna.UID != snb.UID || sna.GID != snb.GID ||
len(sna.Paths) != len(snb.Paths) || len(sna.Excludes) != len(snb.Excludes) ||
len(sna.Tags) != len(snb.Tags) {
return false
}
if !sna.HasPaths(snb.Paths) || !sna.HasTags(snb.Tags) {
return false
}
for i, a := range sna.Excludes {
if a != snb.Excludes[i] {
return false
}
}
return true
}
const numCopyWorkers = 8
func copyTree(ctx context.Context, srcRepo restic.Repository, dstRepo restic.Repository,
visitedTrees restic.IDSet, rootTreeID restic.ID) error {
idChan := make(chan restic.ID)
wg, ctx := errgroup.WithContext(ctx)
treeStream := restic.StreamTrees(ctx, wg, srcRepo, restic.IDs{rootTreeID}, func(treeID restic.ID) bool {
visited := visitedTrees.Has(treeID)
visitedTrees.Insert(treeID)
return visited
}, nil)
wg.Go(func() error {
defer close(idChan)
// reused buffer
var buf []byte
for tree := range treeStream {
if tree.Error != nil {
return fmt.Errorf("LoadTree(%v) returned error %v", tree.ID.Str(), tree.Error)
}
// Do we already have this tree blob?
if !dstRepo.Index().Has(restic.BlobHandle{ID: tree.ID, Type: restic.TreeBlob}) {
// copy raw tree bytes to avoid problems if the serialization changes
var err error
buf, err = srcRepo.LoadBlob(ctx, restic.TreeBlob, tree.ID, buf)
if err != nil {
return fmt.Errorf("LoadBlob(%v) for tree returned error %v", tree.ID, err)
}
_, _, err = dstRepo.SaveBlob(ctx, restic.TreeBlob, buf, tree.ID, false)
if err != nil {
return fmt.Errorf("SaveBlob(%v) for tree returned error %v", tree.ID.Str(), err)
}
}
for _, entry := range tree.Nodes {
// Recursion into directories is handled by StreamTrees
// Copy the blobs for this file.
for _, blobID := range entry.Content {
select {
case idChan <- blobID:
case <-ctx.Done():
return ctx.Err()
}
}
}
}
return nil
})
for i := 0; i < numCopyWorkers; i++ {
wg.Go(func() error {
// reused buffer
var buf []byte
for blobID := range idChan {
// Do we already have this data blob?
if dstRepo.Index().Has(restic.BlobHandle{ID: blobID, Type: restic.DataBlob}) {
continue
}
debug.Log("Copying blob %s\n", blobID.Str())
var err error
buf, err = srcRepo.LoadBlob(ctx, restic.DataBlob, blobID, buf)
if err != nil {
return fmt.Errorf("LoadBlob(%v) returned error %v", blobID, err)
}
_, _, err = dstRepo.SaveBlob(ctx, restic.DataBlob, buf, blobID, false)
if err != nil {
return fmt.Errorf("SaveBlob(%v) returned error %v", blobID, err)
}
}
return nil
})
}
return wg.Wait()
}