1
0
mirror of https://github.com/restic/restic.git synced 2024-07-19 11:17:31 +02:00
restic/internal/index/master_index.go
Michael Eischer 462b82a060 index: reduce size of compressed indexes
use the same index size for compressed and uncompressed indexes.
Otherwise, decoding the index of a compressed repository requires
significantly more memory.
2024-05-24 22:18:14 +02:00

654 lines
16 KiB
Go

package index
import (
"context"
"fmt"
"runtime"
"sync"
"github.com/restic/restic/internal/debug"
"github.com/restic/restic/internal/restic"
"github.com/restic/restic/internal/ui/progress"
"golang.org/x/sync/errgroup"
)
// MasterIndex is a collection of indexes and IDs of chunks that are in the process of being saved.
type MasterIndex struct {
idx []*Index
pendingBlobs restic.BlobSet
idxMutex sync.RWMutex
}
// NewMasterIndex creates a new master index.
func NewMasterIndex() *MasterIndex {
mi := &MasterIndex{pendingBlobs: restic.NewBlobSet()}
mi.clear()
return mi
}
func (mi *MasterIndex) clear() {
// Always add an empty final index, such that MergeFinalIndexes can merge into this.
mi.idx = []*Index{NewIndex()}
mi.idx[0].Finalize()
}
// Lookup queries all known Indexes for the ID and returns all matches.
func (mi *MasterIndex) Lookup(bh restic.BlobHandle) (pbs []restic.PackedBlob) {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
for _, idx := range mi.idx {
pbs = idx.Lookup(bh, pbs)
}
return pbs
}
// LookupSize queries all known Indexes for the ID and returns the first match.
func (mi *MasterIndex) LookupSize(bh restic.BlobHandle) (uint, bool) {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
for _, idx := range mi.idx {
if size, found := idx.LookupSize(bh); found {
return size, found
}
}
return 0, false
}
// AddPending adds a given blob to list of pending Blobs
// Before doing so it checks if this blob is already known.
// Returns true if adding was successful and false if the blob
// was already known
func (mi *MasterIndex) AddPending(bh restic.BlobHandle) bool {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
// Check if blob is pending or in index
if mi.pendingBlobs.Has(bh) {
return false
}
for _, idx := range mi.idx {
if idx.Has(bh) {
return false
}
}
// really not known -> insert
mi.pendingBlobs.Insert(bh)
return true
}
// Has queries all known Indexes for the ID and returns the first match.
// Also returns true if the ID is pending.
func (mi *MasterIndex) Has(bh restic.BlobHandle) bool {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
// also return true if blob is pending
if mi.pendingBlobs.Has(bh) {
return true
}
for _, idx := range mi.idx {
if idx.Has(bh) {
return true
}
}
return false
}
// IDs returns the IDs of all indexes contained in the index.
func (mi *MasterIndex) IDs() restic.IDSet {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
ids := restic.NewIDSet()
for _, idx := range mi.idx {
if !idx.Final() {
continue
}
indexIDs, err := idx.IDs()
if err != nil {
debug.Log("not using index, ID() returned error %v", err)
continue
}
for _, id := range indexIDs {
ids.Insert(id)
}
}
return ids
}
// Packs returns all packs that are covered by the index.
// If packBlacklist is given, those packs are only contained in the
// resulting IDSet if they are contained in a non-final (newly written) index.
func (mi *MasterIndex) Packs(packBlacklist restic.IDSet) restic.IDSet {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
packs := restic.NewIDSet()
for _, idx := range mi.idx {
idxPacks := idx.Packs()
if idx.final && len(packBlacklist) > 0 {
idxPacks = idxPacks.Sub(packBlacklist)
}
packs.Merge(idxPacks)
}
return packs
}
// Insert adds a new index to the MasterIndex.
func (mi *MasterIndex) Insert(idx *Index) {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
mi.idx = append(mi.idx, idx)
}
// StorePack remembers the id and pack in the index.
func (mi *MasterIndex) StorePack(id restic.ID, blobs []restic.Blob) {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
// delete blobs from pending
for _, blob := range blobs {
mi.pendingBlobs.Delete(restic.BlobHandle{Type: blob.Type, ID: blob.ID})
}
for _, idx := range mi.idx {
if !idx.Final() {
idx.StorePack(id, blobs)
return
}
}
newIdx := NewIndex()
newIdx.StorePack(id, blobs)
mi.idx = append(mi.idx, newIdx)
}
// finalizeNotFinalIndexes finalizes all indexes that
// have not yet been saved and returns that list
func (mi *MasterIndex) finalizeNotFinalIndexes() []*Index {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
var list []*Index
for _, idx := range mi.idx {
if !idx.Final() {
idx.Finalize()
list = append(list, idx)
}
}
debug.Log("return %d indexes", len(list))
return list
}
// finalizeFullIndexes finalizes all indexes that are full and returns that list.
func (mi *MasterIndex) finalizeFullIndexes() []*Index {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
var list []*Index
debug.Log("checking %d indexes", len(mi.idx))
for _, idx := range mi.idx {
if idx.Final() {
continue
}
if IndexFull(idx) {
debug.Log("index %p is full", idx)
idx.Finalize()
list = append(list, idx)
} else {
debug.Log("index %p not full", idx)
}
}
debug.Log("return %d indexes", len(list))
return list
}
// Each runs fn on all blobs known to the index. When the context is cancelled,
// the index iteration return immediately. This blocks any modification of the index.
func (mi *MasterIndex) Each(ctx context.Context, fn func(restic.PackedBlob)) error {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
for _, idx := range mi.idx {
if err := idx.Each(ctx, fn); err != nil {
return err
}
}
return nil
}
// MergeFinalIndexes merges all final indexes together.
// After calling, there will be only one big final index in MasterIndex
// containing all final index contents.
// Indexes that are not final are left untouched.
// This merging can only be called after all index files are loaded - as
// removing of superseded index contents is only possible for unmerged indexes.
func (mi *MasterIndex) MergeFinalIndexes() error {
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
// The first index is always final and the one to merge into
newIdx := mi.idx[:1]
for i := 1; i < len(mi.idx); i++ {
idx := mi.idx[i]
// clear reference in masterindex as it may become stale
mi.idx[i] = nil
// do not merge indexes that have no id set
ids, _ := idx.IDs()
if !idx.Final() || len(ids) == 0 {
newIdx = append(newIdx, idx)
} else {
err := mi.idx[0].merge(idx)
if err != nil {
return fmt.Errorf("MergeFinalIndexes: %w", err)
}
}
}
mi.idx = newIdx
return nil
}
func (mi *MasterIndex) Load(ctx context.Context, r restic.ListerLoaderUnpacked, p *progress.Counter, cb func(id restic.ID, idx *Index, oldFormat bool, err error) error) error {
indexList, err := restic.MemorizeList(ctx, r, restic.IndexFile)
if err != nil {
return err
}
if p != nil {
var numIndexFiles uint64
err := indexList.List(ctx, restic.IndexFile, func(_ restic.ID, _ int64) error {
numIndexFiles++
return nil
})
if err != nil {
return err
}
p.SetMax(numIndexFiles)
defer p.Done()
}
err = ForAllIndexes(ctx, indexList, r, func(id restic.ID, idx *Index, oldFormat bool, err error) error {
if p != nil {
p.Add(1)
}
if cb != nil {
err = cb(id, idx, oldFormat, err)
}
if err != nil {
return err
}
// special case to allow check to ignore index loading errors
if idx == nil {
return nil
}
mi.Insert(idx)
return nil
})
if err != nil {
return err
}
return mi.MergeFinalIndexes()
}
type MasterIndexRewriteOpts struct {
SaveProgress *progress.Counter
DeleteProgress func() *progress.Counter
DeleteReport func(id restic.ID, err error)
}
// Rewrite removes packs whose ID is in excludePacks from all known indexes.
// It also removes the rewritten index files and those listed in extraObsolete.
// If oldIndexes is not nil, then only the indexes in this set are processed.
// This is used by repair index to only rewrite and delete the old indexes.
//
// Must not be called concurrently to any other MasterIndex operation.
func (mi *MasterIndex) Rewrite(ctx context.Context, repo restic.Unpacked, excludePacks restic.IDSet, oldIndexes restic.IDSet, extraObsolete restic.IDs, opts MasterIndexRewriteOpts) error {
for _, idx := range mi.idx {
if !idx.Final() {
panic("internal error - index must be saved before calling MasterIndex.Rewrite")
}
}
var indexes restic.IDSet
if oldIndexes != nil {
// repair index adds new index entries for already existing pack files
// only remove the old (possibly broken) entries by only processing old indexes
indexes = oldIndexes
} else {
indexes = mi.IDs()
}
p := opts.SaveProgress
p.SetMax(uint64(len(indexes)))
// reset state which is not necessary for Rewrite and just consumes a lot of memory
// the index state would be invalid after Rewrite completes anyways
mi.clear()
runtime.GC()
// copy excludePacks to prevent unintended sideeffects
excludePacks = excludePacks.Clone()
debug.Log("start rebuilding index of %d indexes, excludePacks: %v", len(indexes), excludePacks)
wg, wgCtx := errgroup.WithContext(ctx)
idxCh := make(chan restic.ID)
wg.Go(func() error {
defer close(idxCh)
for id := range indexes {
select {
case idxCh <- id:
case <-wgCtx.Done():
return wgCtx.Err()
}
}
return nil
})
var rewriteWg sync.WaitGroup
type rewriteTask struct {
idx *Index
oldFormat bool
}
rewriteCh := make(chan rewriteTask)
loader := func() error {
defer rewriteWg.Done()
for id := range idxCh {
buf, err := repo.LoadUnpacked(wgCtx, restic.IndexFile, id)
if err != nil {
return fmt.Errorf("LoadUnpacked(%v): %w", id.Str(), err)
}
idx, oldFormat, err := DecodeIndex(buf, id)
if err != nil {
return err
}
select {
case rewriteCh <- rewriteTask{idx, oldFormat}:
case <-wgCtx.Done():
return wgCtx.Err()
}
}
return nil
}
// loading an index can take quite some time such that this is probably CPU-bound
// the index files are probably already cached at this point
loaderCount := runtime.GOMAXPROCS(0)
// run workers on ch
for i := 0; i < loaderCount; i++ {
rewriteWg.Add(1)
wg.Go(loader)
}
wg.Go(func() error {
rewriteWg.Wait()
close(rewriteCh)
return nil
})
obsolete := restic.NewIDSet(extraObsolete...)
saveCh := make(chan *Index)
wg.Go(func() error {
defer close(saveCh)
newIndex := NewIndex()
for task := range rewriteCh {
// always rewrite indexes using the old format, that include a pack that must be removed or that are not full
if !task.oldFormat && len(task.idx.Packs().Intersect(excludePacks)) == 0 && IndexFull(task.idx) {
// make sure that each pack is only stored exactly once in the index
excludePacks.Merge(task.idx.Packs())
// index is already up to date
p.Add(1)
continue
}
ids, err := task.idx.IDs()
if err != nil || len(ids) != 1 {
panic("internal error, index has no ID")
}
obsolete.Merge(restic.NewIDSet(ids...))
for pbs := range task.idx.EachByPack(wgCtx, excludePacks) {
newIndex.StorePack(pbs.PackID, pbs.Blobs)
if IndexFull(newIndex) {
select {
case saveCh <- newIndex:
case <-wgCtx.Done():
return wgCtx.Err()
}
newIndex = NewIndex()
}
}
if wgCtx.Err() != nil {
return wgCtx.Err()
}
// make sure that each pack is only stored exactly once in the index
excludePacks.Merge(task.idx.Packs())
p.Add(1)
}
select {
case saveCh <- newIndex:
case <-wgCtx.Done():
}
return nil
})
// a worker receives an index from ch, and saves the index
worker := func() error {
for idx := range saveCh {
idx.Finalize()
if _, err := idx.SaveIndex(wgCtx, repo); err != nil {
return err
}
}
return nil
}
// encoding an index can take quite some time such that this can be CPU- or IO-bound
// do not add repo.Connections() here as there are already the loader goroutines.
workerCount := runtime.GOMAXPROCS(0)
// run workers on ch
for i := 0; i < workerCount; i++ {
wg.Go(worker)
}
err := wg.Wait()
p.Done()
if err != nil {
return fmt.Errorf("failed to rewrite indexes: %w", err)
}
p = nil
if opts.DeleteProgress != nil {
p = opts.DeleteProgress()
}
defer p.Done()
return restic.ParallelRemove(ctx, repo, obsolete, restic.IndexFile, func(id restic.ID, err error) error {
if opts.DeleteReport != nil {
opts.DeleteReport(id, err)
}
return err
}, p)
}
// SaveFallback saves all known indexes to index files, leaving out any
// packs whose ID is contained in packBlacklist from finalized indexes.
// It is only intended for use by prune with the UnsafeRecovery option.
//
// Must not be called concurrently to any other MasterIndex operation.
func (mi *MasterIndex) SaveFallback(ctx context.Context, repo restic.SaverRemoverUnpacked, excludePacks restic.IDSet, p *progress.Counter) error {
p.SetMax(uint64(len(mi.Packs(excludePacks))))
mi.idxMutex.Lock()
defer mi.idxMutex.Unlock()
debug.Log("start rebuilding index of %d indexes, excludePacks: %v", len(mi.idx), excludePacks)
obsolete := restic.NewIDSet()
wg, wgCtx := errgroup.WithContext(ctx)
ch := make(chan *Index)
wg.Go(func() error {
defer close(ch)
newIndex := NewIndex()
for _, idx := range mi.idx {
if idx.Final() {
ids, err := idx.IDs()
if err != nil {
panic("internal error - finalized index without ID")
}
debug.Log("adding index ids %v to supersedes field", ids)
obsolete.Merge(restic.NewIDSet(ids...))
}
for pbs := range idx.EachByPack(wgCtx, excludePacks) {
newIndex.StorePack(pbs.PackID, pbs.Blobs)
p.Add(1)
if IndexFull(newIndex) {
select {
case ch <- newIndex:
case <-wgCtx.Done():
return wgCtx.Err()
}
newIndex = NewIndex()
}
}
if wgCtx.Err() != nil {
return wgCtx.Err()
}
}
select {
case ch <- newIndex:
case <-wgCtx.Done():
}
return nil
})
// a worker receives an index from ch, and saves the index
worker := func() error {
for idx := range ch {
idx.Finalize()
if _, err := idx.SaveIndex(wgCtx, repo); err != nil {
return err
}
}
return nil
}
// keep concurrency bounded as we're on a fallback path
workerCount := int(repo.Connections())
// run workers on ch
for i := 0; i < workerCount; i++ {
wg.Go(worker)
}
err := wg.Wait()
p.Done()
// the index no longer matches to stored state
mi.clear()
return err
}
// saveIndex saves all indexes in the backend.
func (mi *MasterIndex) saveIndex(ctx context.Context, r restic.SaverUnpacked, indexes ...*Index) error {
for i, idx := range indexes {
debug.Log("Saving index %d", i)
sid, err := idx.SaveIndex(ctx, r)
if err != nil {
return err
}
debug.Log("Saved index %d as %v", i, sid)
}
return mi.MergeFinalIndexes()
}
// SaveIndex saves all new indexes in the backend.
func (mi *MasterIndex) SaveIndex(ctx context.Context, r restic.SaverUnpacked) error {
return mi.saveIndex(ctx, r, mi.finalizeNotFinalIndexes()...)
}
// SaveFullIndex saves all full indexes in the backend.
func (mi *MasterIndex) SaveFullIndex(ctx context.Context, r restic.SaverUnpacked) error {
return mi.saveIndex(ctx, r, mi.finalizeFullIndexes()...)
}
// ListPacks returns the blobs of the specified pack files grouped by pack file.
func (mi *MasterIndex) ListPacks(ctx context.Context, packs restic.IDSet) <-chan restic.PackBlobs {
out := make(chan restic.PackBlobs)
go func() {
defer close(out)
// only resort a part of the index to keep the memory overhead bounded
for i := byte(0); i < 16; i++ {
packBlob := make(map[restic.ID][]restic.Blob)
for pack := range packs {
if pack[0]&0xf == i {
packBlob[pack] = nil
}
}
if len(packBlob) == 0 {
continue
}
err := mi.Each(ctx, func(pb restic.PackedBlob) {
if packs.Has(pb.PackID) && pb.PackID[0]&0xf == i {
packBlob[pb.PackID] = append(packBlob[pb.PackID], pb.Blob)
}
})
if err != nil {
return
}
// pass on packs
for packID, pbs := range packBlob {
// allow GC
packBlob[packID] = nil
select {
case out <- restic.PackBlobs{PackID: packID, Blobs: pbs}:
case <-ctx.Done():
return
}
}
}
}()
return out
}
// Only for use by AssociatedSet
func (mi *MasterIndex) blobIndex(h restic.BlobHandle) int {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
// other indexes are ignored as their ids can change when merged into the main index
return mi.idx[0].BlobIndex(h)
}
// Only for use by AssociatedSet
func (mi *MasterIndex) stableLen(t restic.BlobType) uint {
mi.idxMutex.RLock()
defer mi.idxMutex.RUnlock()
// other indexes are ignored as their ids can change when merged into the main index
return mi.idx[0].Len(t)
}