core/state/snapshot: use local bloom filters

core/state/snapshot/difflayer.go

@@ -27,7 +27,7 @@ import (
 
 	"github.com/ethereum/go-ethereum/common"
 	"github.com/ethereum/go-ethereum/rlp"
-	"github.com/steakknife/bloomfilter"
+	"github.com/holiman/bloomfilter"
 )
 
 var (
@@ -110,7 +110,8 @@ type diffLayer struct {
 	storageList map[common.Hash][]common.Hash          // List of storage slots for iterated retrievals, one per account. Any existing lists are sorted if non-nil
 	storageData map[common.Hash]map[common.Hash][]byte // Keyed storage slots for direct retrival. one per account (nil means deleted)
 
-	diffed *bloomfilter.Filter // Bloom filter tracking all the diffed items up to the disk layer
+	diffed     *bloomfilter.Filter // Bloom filter tracking local diffed items
+	cumulative *bloomfilter.Filter // cumulumative bloom filter -- by default set to nil
 
 	lock sync.RWMutex
 }
@@ -169,34 +170,24 @@ func newDiffLayer(parent snapshot, root common.Hash, destructs map[common.Hash]s
 		accountData: accounts,
 		storageData: storage,
 	}
-	switch parent := parent.(type) {
+	dl.initBloom()
-	case *diskLayer:
-		dl.rebloom(parent, true)
-	case *diffLayer:
-		dl.rebloom(parent.origin, true)
-	default:
-		panic("unknown parent type")
-	}
 	return dl
 }
 
-// rebloom discards the layer's current bloom and rebuilds it from scratch based
+// initBloom builds the layer's bloom and calculates memory consumption
-// on the parent's and the local diffs.
+func (dl *diffLayer) initBloom() {
-func (dl *diffLayer) rebloom(origin *diskLayer, creation bool) {
 	dl.lock.Lock()
 	defer dl.lock.Unlock()
 
 	defer func(start time.Time) {
 		snapshotBloomIndexTimer.Update(time.Since(start))
 	}(time.Now())
-
-	// Inject the new origin that triggered the rebloom
-	dl.origin = origin
-
 	// Retrieve the parent bloom or create a fresh empty one
+	// We need to use the same keys as the parent bloom if we want to be able
+	// to make union later
 	if parent, ok := dl.parent.(*diffLayer); ok {
 		parent.lock.RLock()
-		dl.diffed, _ = parent.diffed.Copy()
+		dl.diffed, _ = parent.diffed.NewCompatible()
 		parent.lock.RUnlock()
 	} else {
 		dl.diffed, _ = bloomfilter.New(uint64(bloomSize), uint64(bloomFuncs))
@@ -218,9 +209,7 @@ func (dl *diffLayer) rebloom(origin *diskLayer, creation bool) {
 		nHashes++
 	}
 	dl.memory = dataSize + nHashes*uint64(common.HashLength)
-	if creation {
+	snapshotDirtyAccountWriteMeter.Mark(int64(dataSize))
-		snapshotDirtyAccountWriteMeter.Mark(int64(dataSize))
-	}
 
 	dataSize, nHashes = uint64(0), uint64(0)
 	for accountHash, slots := range dl.storageData {
@@ -235,16 +224,39 @@ func (dl *diffLayer) rebloom(origin *diskLayer, creation bool) {
 		}
 	}
 	dl.memory += dataSize + nHashes*uint64(common.HashLength)
-	if creation {
+	snapshotDirtyStorageWriteMeter.Mark(int64(dataSize))
-		snapshotDirtyStorageWriteMeter.Mark(int64(dataSize))
+}
+
+// Prepare prepares the difflayer for execution, and creates the cumulative
+// bloom
+func (dl *diffLayer) Prepare(origin *diskLayer) {
+	dl.lock.Lock()
+	dl.cumulative, _ = dl.diffed.Copy()
+	layer := dl
+	for {
+		if parent, ok := layer.parent.(*diffLayer); ok {
+			parent.lock.RLock()
+			dl.cumulative.UnionInPlace(parent.diffed)
+			parent.lock.RUnlock()
+			layer = parent
+		} else {
+			break
+		}
 	}
+	dl.origin = origin
+	dl.lock.Unlock()
 	// Calculate the current false positive rate and update the error rate meter.
-	// This is a bit cheating because subsequent layers will overwrite it, but it
+	k := float64(dl.cumulative.K())
-	// should be fine, we're only interested in ballpark figures.
+	n := float64(dl.cumulative.N())
-	k := float64(dl.diffed.K())
+	m := float64(dl.cumulative.M())
-	n := float64(dl.diffed.N())
-	m := float64(dl.diffed.M())
 	snapshotBloomErrorGauge.Update(math.Pow(1.0-math.Exp((-k)*(n+0.5)/(m-1)), k))
+
+}
+
+func (dl *diffLayer) Release() {
+	dl.lock.Lock()
+	dl.cumulative = nil
+	dl.lock.Unlock()
 }
 
 // Root returns the root hash for which this snapshot was made.
@@ -286,9 +298,9 @@ func (dl *diffLayer) AccountRLP(hash common.Hash) ([]byte, error) {
 	// Check the bloom filter first whether there's even a point in reaching into
 	// all the maps in all the layers below
 	dl.lock.RLock()
-	hit := dl.diffed.Contains(accountBloomHasher(hash))
+	hit := dl.cumulative.Contains(accountBloomHasher(hash))
 	if !hit {
-		hit = dl.diffed.Contains(destructBloomHasher(hash))
+		hit = dl.cumulative.Contains(destructBloomHasher(hash))
 	}
 	dl.lock.RUnlock()
 
@@ -346,9 +358,9 @@ func (dl *diffLayer) Storage(accountHash, storageHash common.Hash) ([]byte, erro
 	// Check the bloom filter first whether there's even a point in reaching into
 	// all the maps in all the layers below
 	dl.lock.RLock()
-	hit := dl.diffed.Contains(storageBloomHasher{accountHash, storageHash})
+	hit := dl.cumulative.Contains(storageBloomHasher{accountHash, storageHash})
 	if !hit {
-		hit = dl.diffed.Contains(destructBloomHasher(accountHash))
+		hit = dl.cumulative.Contains(destructBloomHasher(accountHash))
 	}
 	dl.lock.RUnlock()
 
@@ -457,6 +469,7 @@ func (dl *diffLayer) flatten() snapshot {
 		parent.storageData[accountHash] = comboData
 	}
 	// Return the combo parent
+	parent.diffed.UnionInPlace(dl.diffed)
 	return &diffLayer{
 		parent:      parent.parent,
 		origin:      parent.origin,
@@ -465,7 +478,7 @@ func (dl *diffLayer) flatten() snapshot {
 		accountData: parent.accountData,
 		storageData: parent.storageData,
 		storageList: make(map[common.Hash][]common.Hash),
-		diffed:      dl.diffed,
+		diffed:      parent.diffed,
 		memory:      parent.memory + dl.memory,
 	}
 }

core/state/snapshot/disklayer.go

@@ -164,3 +164,6 @@ func (dl *diskLayer) Storage(accountHash, storageHash common.Hash) ([]byte, erro
 func (dl *diskLayer) Update(blockHash common.Hash, destructs map[common.Hash]struct{}, accounts map[common.Hash][]byte, storage map[common.Hash]map[common.Hash][]byte) *diffLayer {
 	return newDiffLayer(dl, blockHash, destructs, accounts, storage)
 }
+
+func (dl *diskLayer) Prepare(*diskLayer) {}
+func (dl *diskLayer) Release()           {}

core/state/snapshot/snapshot.go

@@ -107,6 +107,7 @@ type Snapshot interface {
 	// Storage directly retrieves the storage data associated with a particular hash,
 	// within a particular account.
 	Storage(accountHash, storageHash common.Hash) ([]byte, error)
+	Release()
 }
 
 // snapshot is the internal version of the snapshot data layer that supports some
@@ -138,6 +139,8 @@ type snapshot interface {
 
 	// AccountIterator creates an account iterator over an arbitrary layer.
 	AccountIterator(seek common.Hash) AccountIterator
+
+	Prepare(*diskLayer)
 }
 
 // SnapshotTree is an Ethereum state snapshot tree. It consists of one persistent
@@ -150,11 +153,12 @@ type snapshot interface {
 // storage data to avoid expensive multi-level trie lookups; and to allow sorted,
 // cheap iteration of the account/storage tries for sync aid.
 type Tree struct {
-	diskdb ethdb.KeyValueStore      // Persistent database to store the snapshot
+	diskdb    ethdb.KeyValueStore      // Persistent database to store the snapshot
-	triedb *trie.Database           // In-memory cache to access the trie through
+	triedb    *trie.Database           // In-memory cache to access the trie through
-	cache  int                      // Megabytes permitted to use for read caches
+	cache     int                      // Megabytes permitted to use for read caches
-	layers map[common.Hash]snapshot // Collection of all known layers
+	layers    map[common.Hash]snapshot // Collection of all known layers
-	lock   sync.RWMutex
+	lock      sync.RWMutex
+	diskLayer *diskLayer // The underlying disklayer
 }
 
 // New attempts to load an already existing snapshot from a persistent key-value
@@ -186,6 +190,9 @@ func New(diskdb ethdb.KeyValueStore, triedb *trie.Database, cache int, root comm
 	for head != nil {
 		snap.layers[head.Root()] = head
 		head = head.Parent()
+		// TODO(@holiman), check if we need to add this back:
+		//case *diskLayer:
+		//	snap.diskLayer = self
 	}
 	return snap
 }
@@ -211,6 +218,16 @@ func (t *Tree) waitBuild() {
 	}
 }
 
+func (t *Tree) PrepareSnapshot(blockRoot common.Hash) Snapshot {
+	t.lock.RLock()
+	defer t.lock.RUnlock()
+	if snap := t.layers[blockRoot]; snap != nil {
+		snap.Prepare(t.diskLayer)
+		return snap
+	}
+	return nil
+}
+
 // Snapshot retrieves a snapshot belonging to the given block root, or nil if no
 // snapshot is maintained for that block.
 func (t *Tree) Snapshot(blockRoot common.Hash) Snapshot {
@@ -278,6 +295,7 @@ func (t *Tree) Cap(root common.Hash, layers int) error {
 
 		// Replace the entire snapshot tree with the flat base
 		t.layers = map[common.Hash]snapshot{base.root: base}
+		t.diskLayer = base
 		return nil
 
 	case 1:
@@ -291,6 +309,7 @@ func (t *Tree) Cap(root common.Hash, layers int) error {
 		bottom = diff.flatten().(*diffLayer)
 		if bottom.memory >= aggregatorMemoryLimit {
 			base = diffToDisk(bottom)
+			t.diskLayer = base
 		}
 		diff.lock.RUnlock()
 
@@ -305,6 +324,9 @@ func (t *Tree) Cap(root common.Hash, layers int) error {
 	default:
 		// Many layers requested to be retained, cap normally
 		persisted = t.cap(diff, layers)
+		if persisted != nil {
+			t.diskLayer = persisted
+		}
 	}
 	// Remove any layer that is stale or links into a stale layer
 	children := make(map[common.Hash][]common.Hash)
@@ -328,18 +350,6 @@ func (t *Tree) Cap(root common.Hash, layers int) error {
 		}
 	}
 	// If the disk layer was modified, regenerate all the cummulative blooms
-	if persisted != nil {
-		var rebloom func(root common.Hash)
-		rebloom = func(root common.Hash) {
-			if diff, ok := t.layers[root].(*diffLayer); ok {
-				diff.rebloom(persisted, false)
-			}
-			for _, child := range children[root] {
-				rebloom(child)
-			}
-		}
-		rebloom(persisted.root)
-	}
 	return nil
 }
 
@@ -591,8 +601,10 @@ func (t *Tree) Rebuild(root common.Hash) {
 	// Start generating a new snapshot from scratch on a backgroung thread. The
 	// generator will run a wiper first if there's not one running right now.
 	log.Info("Rebuilding state snapshot")
+	diskLayer := generateSnapshot(t.diskdb, t.triedb, t.cache, root, wiper)
+	t.diskLayer = diskLayer
 	t.layers = map[common.Hash]snapshot{
-		root: generateSnapshot(t.diskdb, t.triedb, t.cache, root, wiper),
+		root: diskLayer,
 	}
 }
 

core/state/statedb.go

@@ -133,7 +133,7 @@ func New(root common.Hash, db Database, snaps *snapshot.Tree) (*StateDB, error)
 		journal:             newJournal(),
 	}
 	if sdb.snaps != nil {
-		if sdb.snap = sdb.snaps.Snapshot(root); sdb.snap != nil {
+		if sdb.snap = sdb.snaps.PrepareSnapshot(root); sdb.snap != nil {
 			sdb.snapDestructs = make(map[common.Hash]struct{})
 			sdb.snapAccounts = make(map[common.Hash][]byte)
 			sdb.snapStorage = make(map[common.Hash]map[common.Hash][]byte)
@@ -862,7 +862,11 @@ func (s *StateDB) Commit(deleteEmptyObjects bool) (common.Hash, error) {
 			if err := s.snaps.Update(root, parent, s.snapDestructs, s.snapAccounts, s.snapStorage); err != nil {
 				log.Warn("Failed to update snapshot tree", "from", parent, "to", root, "err", err)
 			}
+			if err := s.snaps.Cap(root, 127); err != nil { // Persistent layer is 128th, the last available trie
+				log.Warn("Failed to cap snapshot tree", "root", root, "layers", 127, "err", err)
+			}
 		}
+		s.snap.Release()
 		s.snap, s.snapDestructs, s.snapAccounts, s.snapStorage = nil, nil, nil, nil
 	}
 	return root, err

go.mod

@@ -33,6 +33,7 @@ require (
 	github.com/gorilla/websocket v1.4.1-0.20190629185528-ae1634f6a989
 	github.com/graph-gophers/graphql-go v0.0.0-20191115155744-f33e81362277
 	github.com/hashicorp/golang-lru v0.0.0-20160813221303-0a025b7e63ad
+	github.com/holiman/bloomfilter v0.0.0-20191204204232-5fc8e9365fc2
 	github.com/huin/goupnp v0.0.0-20161224104101-679507af18f3
 	github.com/influxdata/influxdb v1.2.3-0.20180221223340-01288bdb0883
 	github.com/jackpal/go-nat-pmp v1.0.2-0.20160603034137-1fa385a6f458
@@ -57,6 +58,7 @@ require (
 	github.com/stretchr/testify v1.4.0
 	github.com/syndtr/goleveldb v1.0.1-0.20190923125748-758128399b1d
 	github.com/tyler-smith/go-bip39 v1.0.1-0.20181017060643-dbb3b84ba2ef
+	github.com/umbracle/fastrlp v0.0.0-20191017143648-86584926e68c
 	github.com/wsddn/go-ecdh v0.0.0-20161211032359-48726bab9208
 	golang.org/x/crypto v0.0.0-20200311171314-f7b00557c8c4
 	golang.org/x/net v0.0.0-20200301022130-244492dfa37a // indirect

go.sum

@@ -99,6 +99,8 @@ github.com/graph-gophers/graphql-go v0.0.0-20191115155744-f33e81362277 h1:E0whKx
 github.com/graph-gophers/graphql-go v0.0.0-20191115155744-f33e81362277/go.mod h1:9CQHMSxwO4MprSdzoIEobiHpoLtHm77vfxsvsIN5Vuc=
 github.com/hashicorp/golang-lru v0.0.0-20160813221303-0a025b7e63ad h1:eMxs9EL0PvIGS9TTtxg4R+JxuPGav82J8rA+GFnY7po=
 github.com/hashicorp/golang-lru v0.0.0-20160813221303-0a025b7e63ad/go.mod h1:/m3WP610KZHVQ1SGc6re/UDhFvYD7pJ4Ao+sR/qLZy8=
+github.com/holiman/bloomfilter v0.0.0-20191204204232-5fc8e9365fc2 h1:W++/MG1xoR0V8YM2H47YC/G/DnZKODb0TmxY/GBcEac=
+github.com/holiman/bloomfilter v0.0.0-20191204204232-5fc8e9365fc2/go.mod h1:+E92jn6hSglI3YfM9VLe+w2tHKtSL9l83iBiZS7/4Lw=
 github.com/hpcloud/tail v1.0.0 h1:nfCOvKYfkgYP8hkirhJocXT2+zOD8yUNjXaWfTlyFKI=
 github.com/hpcloud/tail v1.0.0/go.mod h1:ab1qPbhIpdTxEkNHXyeSf5vhxWSCs/tWer42PpOxQnU=
 github.com/huin/goupnp v0.0.0-20161224104101-679507af18f3 h1:DqD8eigqlUm0+znmx7zhL0xvTW3+e1jCekJMfBUADWI=
@@ -189,6 +191,8 @@ github.com/syndtr/goleveldb v1.0.1-0.20190923125748-758128399b1d h1:gZZadD8H+fF+
 github.com/syndtr/goleveldb v1.0.1-0.20190923125748-758128399b1d/go.mod h1:9OrXJhf154huy1nPWmuSrkgjPUtUNhA+Zmy+6AESzuA=
 github.com/tyler-smith/go-bip39 v1.0.1-0.20181017060643-dbb3b84ba2ef h1:wHSqTBrZW24CsNJDfeh9Ex6Pm0Rcpc7qrgKBiL44vF4=
 github.com/tyler-smith/go-bip39 v1.0.1-0.20181017060643-dbb3b84ba2ef/go.mod h1:sJ5fKU0s6JVwZjjcUEX2zFOnvq0ASQ2K9Zr6cf67kNs=
+github.com/umbracle/fastrlp v0.0.0-20191017143648-86584926e68c h1:KCbyOp0afQzDTGARP+qj9HpZVk79WQaQtAO+4kMy/E4=
+github.com/umbracle/fastrlp v0.0.0-20191017143648-86584926e68c/go.mod h1:TBFAIaLMusGh1X7Pd3trJTjLsOmNErpPV9/J7BY9lSA=
 github.com/urfave/cli v1.22.1/go.mod h1:Gos4lmkARVdJ6EkW0WaNv/tZAAMe9V7XWyB60NtXRu0=
 github.com/wsddn/go-ecdh v0.0.0-20161211032359-48726bab9208 h1:1cngl9mPEoITZG8s8cVcUy5CeIBYhEESkOB7m6Gmkrk=
 github.com/wsddn/go-ecdh v0.0.0-20161211032359-48726bab9208/go.mod h1:IotVbo4F+mw0EzQ08zFqg7pK3FebNXpaMsRy2RT+Ees=