Skip to content

Commit

Permalink
revert pruning hotfix
Browse files Browse the repository at this point in the history
  • Loading branch information
@temaniarpit27
temaniarpit27 committed Dec 27, 2023
1 parent 46bd7d7 commit 7253bab
Show file tree
Hide file tree
Showing 2 changed files with 5 additions and 93 deletions.
96 changes: 4 additions & 92 deletions core/state/pruner/pruner.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -57,7 +57,6 @@ const (
// Config includes all the configurations for pruning.
type Config struct {
Datadir string // The directory of the state database
Cachedir string // The directory of state clean cache
BloomSize uint64 // The Megabytes of memory allocated to bloom-filter
}

Expand Down Expand Up @@ -86,14 +85,12 @@ func NewPruner(db ethdb.Database, config Config) (*Pruner, error) {
if headBlock == nil {
return nil, errors.New("failed to load head block")
}

snapconfig := snapshot.Config{
CacheSize: 256,
Recovery: false,
NoBuild: true,
AsyncBuild: false,
}

snaptree, err := snapshot.New(snapconfig, db, trie.NewDatabase(db), headBlock.Root())
if err != nil {
return nil, err // The relevant snapshot(s) might not exist
Expand All @@ -103,12 +100,10 @@ func NewPruner(db ethdb.Database, config Config) (*Pruner, error) {
log.Warn("Sanitizing bloomfilter size", "provided(MB)", config.BloomSize, "updated(MB)", 256)
config.BloomSize = 256
}

stateBloom, err := newStateBloomWithSize(config.BloomSize)
if err != nil {
return nil, err
}

return &Pruner{
config: config,
chainHeader: headBlock.Header(),
Expand All @@ -134,7 +129,6 @@ func prune(snaptree *snapshot.Tree, root common.Hash, maindb ethdb.Database, sta
batch = maindb.NewBatch()
iter = maindb.NewIterator(nil, nil)
)

for iter.Next() {
key := iter.Key()

Expand All @@ -148,34 +142,28 @@ func prune(snaptree *snapshot.Tree, root common.Hash, maindb ethdb.Database, sta
if isCode {
checkKey = codeKey
}

if _, exist := middleStateRoots[common.BytesToHash(checkKey)]; exist {
log.Debug("Forcibly delete the middle state roots", "hash", common.BytesToHash(checkKey))
} else {
if stateBloom.Contain(checkKey) {
continue
}
}

count += 1
size += common.StorageSize(len(key) + len(iter.Value()))
batch.Delete(key)

var eta time.Duration // Realistically will never remain uninited

if done := binary.BigEndian.Uint64(key[:8]); done > 0 {
var (
left = math.MaxUint64 - binary.BigEndian.Uint64(key[:8])
speed = done/uint64(time.Since(pstart)/time.Millisecond+1) + 1 // +1s to avoid division by zero
)

eta = time.Duration(left/speed) * time.Millisecond
}

if time.Since(logged) > 8*time.Second {
log.Info("Pruning state data", "nodes", count, "size", size,
"elapsed", common.PrettyDuration(time.Since(pstart)), "eta", common.PrettyDuration(eta))

logged = time.Now()
}
// Recreate the iterator after every batch commit in order
Expand All @@ -189,12 +177,10 @@ func prune(snaptree *snapshot.Tree, root common.Hash, maindb ethdb.Database, sta
}
}
}

if batch.ValueSize() > 0 {
batch.Write()
batch.Reset()
}

iter.Release()
log.Info("Pruned state data", "nodes", count, "size", size, "elapsed", common.PrettyDuration(time.Since(pstart)))

Expand All @@ -221,36 +207,29 @@ func prune(snaptree *snapshot.Tree, root common.Hash, maindb ethdb.Database, sta
// Note for small pruning, the compaction is skipped.
if count >= rangeCompactionThreshold {
cstart := time.Now()

for b := 0x00; b <= 0xf0; b += 0x10 {
var (
start = []byte{byte(b)}
end = []byte{byte(b + 0x10)}
)

if b == 0xf0 {
end = nil
}

log.Info("Compacting database", "range", fmt.Sprintf("%#x-%#x", start, end), "elapsed", common.PrettyDuration(time.Since(cstart)))

if err := maindb.Compact(start, end); err != nil {
log.Error("Database compaction failed", "error", err)
return err
}
}
log.Info("Database compaction finished", "elapsed", common.PrettyDuration(time.Since(cstart)))
}

log.Info("State pruning successful", "pruned", size, "elapsed", common.PrettyDuration(time.Since(start)))

return nil
}

// Prune deletes all historical state nodes except the nodes belong to the
// specified state version. If user doesn't specify the state version, use
// the bottom-most snapshot diff layer as the target.
// nolint:nestif
func (p *Pruner) Prune(root common.Hash) error {
// If the state bloom filter is already committed previously,
// reuse it for pruning instead of generating a new one. It's
Expand All @@ -260,9 +239,8 @@ func (p *Pruner) Prune(root common.Hash) error {
if err != nil {
return err
}

if stateBloomRoot != (common.Hash{}) {
return RecoverPruning(p.config.Datadir, p.db, p.config.Cachedir)
return RecoverPruning(p.config.Datadir, p.db)
}
// If the target state root is not specified, use the HEAD-127 as the
// target. The reason for picking it is:
Expand All @@ -287,8 +265,8 @@ func (p *Pruner) Prune(root common.Hash) error {
// is the presence of root can indicate the presence of the
// entire trie.
if !rawdb.HasLegacyTrieNode(p.db, root) {
// The special case is for clique based networks(goerli and
// some other private networks), it's possible that two
// The special case is for clique based networks(goerli
// and some other private networks), it's possible that two
// consecutive blocks will have same root. In this case snapshot
// difflayer won't be created. So HEAD-127 may not paired with
// head-127 layer. Instead the paired layer is higher than the
Expand All @@ -299,23 +277,18 @@ func (p *Pruner) Prune(root common.Hash) error {
// state available, but we don't want to use the topmost state
// as the pruning target.
var found bool

for i := len(layers) - 2; i >= 2; i-- {
if rawdb.HasLegacyTrieNode(p.db, layers[i].Root()) {
root = layers[i].Root()
found = true

log.Info("Selecting middle-layer as the pruning target", "root", root, "depth", i)

break
}
}

if !found {
if len(layers) > 0 {
return errors.New("no snapshot paired state")
}

return fmt.Errorf("associated state[%x] is not present", root)
}
} else {
Expand All @@ -325,27 +298,18 @@ func (p *Pruner) Prune(root common.Hash) error {
log.Info("Selecting user-specified state as the pruning target", "root", root)
}
}
// Before start the pruning, delete the clean trie cache first.
// It's necessary otherwise in the next restart we will hit the
// deleted state root in the "clean cache" so that the incomplete
// state is picked for usage.
deleteCleanTrieCache(p.config.Cachedir)

// All the state roots of the middle layer should be forcibly pruned,
// otherwise the dangling state will be left.
middleRoots := make(map[common.Hash]struct{})

for _, layer := range layers {
if layer.Root() == root {
break
}

middleRoots[layer.Root()] = struct{}{}
}
// Traverse the target state, re-construct the whole state trie and
// commit to the given bloom filter.
start := time.Now()

if err := snapshot.GenerateTrie(p.snaptree, root, p.db, p.stateBloom); err != nil {
return err
}
Expand All @@ -354,17 +318,13 @@ func (p *Pruner) Prune(root common.Hash) error {
if err := extractGenesis(p.db, p.stateBloom); err != nil {
return err
}

filterName := bloomFilterName(p.config.Datadir, root)

log.Info("Writing state bloom to disk", "name", filterName)

if err := p.stateBloom.Commit(filterName, filterName+stateBloomFileTempSuffix); err != nil {
return err
}

log.Info("State bloom filter committed", "name", filterName)

return prune(p.snaptree, root, p.db, p.stateBloom, filterName, middleRoots, start)
}

Expand All @@ -375,16 +335,14 @@ func (p *Pruner) Prune(root common.Hash) error {
// pruning can be resumed. What's more if the bloom filter is constructed, the
// pruning **has to be resumed**. Otherwise a lot of dangling nodes may be left
// in the disk.
func RecoverPruning(datadir string, db ethdb.Database, trieCachePath string) error {
func RecoverPruning(datadir string, db ethdb.Database) error {
stateBloomPath, stateBloomRoot, err := findBloomFilter(datadir)
if err != nil {
return err
}

if stateBloomPath == "" {
return nil // nothing to recover
}

headBlock := rawdb.ReadHeadBlock(db)
if headBlock == nil {
return errors.New("failed to load head block")
Expand All @@ -403,47 +361,34 @@ func RecoverPruning(datadir string, db ethdb.Database, trieCachePath string) err
NoBuild: true,
AsyncBuild: false,
}

snaptree, err := snapshot.New(snapconfig, db, trie.NewDatabase(db), headBlock.Root())
if err != nil {
return err // The relevant snapshot(s) might not exist
}

stateBloom, err := NewStateBloomFromDisk(stateBloomPath)
if err != nil {
return err
}

log.Info("Loaded state bloom filter", "path", stateBloomPath)

// Before start the pruning, delete the clean trie cache first.
// It's necessary otherwise in the next restart we will hit the
// deleted state root in the "clean cache" so that the incomplete
// state is picked for usage.
deleteCleanTrieCache(trieCachePath)

// All the state roots of the middle layers should be forcibly pruned,
// otherwise the dangling state will be left.
var (
found bool
layers = snaptree.Snapshots(headBlock.Root(), 128, true)
middleRoots = make(map[common.Hash]struct{})
)

for _, layer := range layers {
if layer.Root() == stateBloomRoot {
found = true
break
}

middleRoots[layer.Root()] = struct{}{}
}

if !found {
log.Error("Pruning target state is not existent")
return errors.New("non-existent target state")
}

return prune(snaptree, stateBloomRoot, db, stateBloom, stateBloomPath, middleRoots, time.Now())
}

Expand All @@ -454,17 +399,14 @@ func extractGenesis(db ethdb.Database, stateBloom *stateBloom) error {
if genesisHash == (common.Hash{}) {
return errors.New("missing genesis hash")
}

genesis := rawdb.ReadBlock(db, genesisHash, 0)
if genesis == nil {
return errors.New("missing genesis block")
}

t, err := trie.NewStateTrie(trie.StateTrieID(genesis.Root()), trie.NewDatabase(db))
if err != nil {
return err
}

accIter, err := t.NodeIterator(nil)
if err != nil {
return err
Expand All @@ -483,15 +425,12 @@ func extractGenesis(db ethdb.Database, stateBloom *stateBloom) error {
if err := rlp.DecodeBytes(accIter.LeafBlob(), &acc); err != nil {
return err
}

if acc.Root != types.EmptyRootHash {
id := trie.StorageTrieID(genesis.Root(), common.BytesToHash(accIter.LeafKey()), acc.Root)

storageTrie, err := trie.NewStateTrie(id, trie.NewDatabase(db))
if err != nil {
return err
}

storageIter, err := storageTrie.NodeIterator(nil)
if err != nil {
return err
Expand All @@ -502,18 +441,15 @@ func extractGenesis(db ethdb.Database, stateBloom *stateBloom) error {
stateBloom.Put(hash.Bytes(), nil)
}
}

if storageIter.Error() != nil {
return storageIter.Error()
}
}

if !bytes.Equal(acc.CodeHash, types.EmptyCodeHash.Bytes()) {
stateBloom.Put(acc.CodeHash, nil)
}
}
}

return accIter.Error()
}

Expand All @@ -526,7 +462,6 @@ func isBloomFilter(filename string) (bool, common.Hash) {
if strings.HasPrefix(filename, stateBloomFilePrefix) && strings.HasSuffix(filename, stateBloomFileSuffix) {
return true, common.HexToHash(filename[len(stateBloomFilePrefix)+1 : len(filename)-len(stateBloomFileSuffix)-1])
}

return false, common.Hash{}
}

Expand All @@ -535,7 +470,6 @@ func findBloomFilter(datadir string) (string, common.Hash, error) {
stateBloomPath string
stateBloomRoot common.Hash
)

if err := filepath.Walk(datadir, func(path string, info os.FileInfo, err error) error {
if info != nil && !info.IsDir() {
ok, root := isBloomFilter(path)
Expand All @@ -548,27 +482,5 @@ func findBloomFilter(datadir string) (string, common.Hash, error) {
}); err != nil {
return "", common.Hash{}, err
}

return stateBloomPath, stateBloomRoot, nil
}

const warningLog = `
WARNING!
The clean trie cache is not found. Please delete it by yourself after the
pruning. Remember don't start the Geth without deleting the clean trie cache
otherwise the entire database may be damaged!
Check the command description "geth snapshot prune-state --help" for more details.
`

func deleteCleanTrieCache(path string) {
if !common.FileExist(path) {
log.Warn(warningLog)
return
}

os.RemoveAll(path)
log.Info("Deleted trie clean cache", "path", path)
}
2 changes: 1 addition & 1 deletion eth/backend.go
View file
Original file line number Diff line number Diff line change
Expand Up @@ -143,7 +143,7 @@ func New(stack *node.Node, config *ethconfig.Config) (*Ethereum, error) {
if err != nil {
return nil, err
}
if err := pruner.RecoverPruning(stack.ResolvePath(""), chainDb, ""); err != nil {
if err := pruner.RecoverPruning(stack.ResolvePath(""), chainDb); err != nil {

Check warning on line 146 in eth/backend.go

View check run for this annotation

Codecov / codecov/patch

eth/backend.go#L146 

Added line #L146 was not covered by tests
log.Error("Failed to recover state", "error", err)
}

Expand Down

0 comments on commit 7253bab

Please sign in to comment.