Merge pull request ethereum#18087 from karalabe/trie-read-cacher

cmd, core, eth, light, trie: add trie read caching layer

cmd/geth/main.go

@@ -89,6 +89,7 @@ var (
 		utils.LightKDFFlag,
 		utils.CacheFlag,
 		utils.CacheDatabaseFlag,
+		utils.CacheTrieFlag,
 		utils.CacheGCFlag,
 		utils.TrieCacheGenFlag,
 		utils.ListenPortFlag,

cmd/geth/usage.go

@@ -132,6 +132,7 @@ var AppHelpFlagGroups = []flagGroup{
 		Flags: []cli.Flag{
 			utils.CacheFlag,
 			utils.CacheDatabaseFlag,
+			utils.CacheTrieFlag,
 			utils.CacheGCFlag,
 			utils.TrieCacheGenFlag,
 		},

cmd/utils/flags.go

@@ -295,7 +295,12 @@ var (
 	CacheDatabaseFlag = cli.IntFlag{
 		Name:  "cache.database",
 		Usage: "Percentage of cache memory allowance to use for database io",
-		Value: 75,
+		Value: 50,
+	}
+	CacheTrieFlag = cli.IntFlag{
+		Name:  "cache.trie",
+		Usage: "Percentage of cache memory allowance to use for trie caching",
+		Value: 25,
 	}
 	CacheGCFlag = cli.IntFlag{
 		Name:  "cache.gc",
@@ -1157,8 +1162,11 @@ func SetEthConfig(ctx *cli.Context, stack *node.Node, cfg *eth.Config) {
 	}
 	cfg.NoPruning = ctx.GlobalString(GCModeFlag.Name) == "archive"
 
+	if ctx.GlobalIsSet(CacheFlag.Name) || ctx.GlobalIsSet(CacheTrieFlag.Name) {
+		cfg.TrieCleanCache = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheTrieFlag.Name) / 100
+	}
 	if ctx.GlobalIsSet(CacheFlag.Name) || ctx.GlobalIsSet(CacheGCFlag.Name) {
-		cfg.TrieCache = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheGCFlag.Name) / 100
+		cfg.TrieDirtyCache = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheGCFlag.Name) / 100
 	}
 	if ctx.GlobalIsSet(MinerNotifyFlag.Name) {
 		cfg.MinerNotify = strings.Split(ctx.GlobalString(MinerNotifyFlag.Name), ",")
@@ -1393,12 +1401,16 @@ func MakeChain(ctx *cli.Context, stack *node.Node) (chain *core.BlockChain, chai
 		Fatalf("--%s must be either 'full' or 'archive'", GCModeFlag.Name)
 	}
 	cache := &core.CacheConfig{
-		Disabled:      ctx.GlobalString(GCModeFlag.Name) == "archive",
-		TrieNodeLimit: eth.DefaultConfig.TrieCache,
-		TrieTimeLimit: eth.DefaultConfig.TrieTimeout,
+		Disabled:       ctx.GlobalString(GCModeFlag.Name) == "archive",
+		TrieCleanLimit: eth.DefaultConfig.TrieCleanCache,
+		TrieDirtyLimit: eth.DefaultConfig.TrieDirtyCache,
+		TrieTimeLimit:  eth.DefaultConfig.TrieTimeout,
+	}
+	if ctx.GlobalIsSet(CacheFlag.Name) || ctx.GlobalIsSet(CacheTrieFlag.Name) {
+		cache.TrieCleanLimit = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheTrieFlag.Name) / 100
 	}
 	if ctx.GlobalIsSet(CacheFlag.Name) || ctx.GlobalIsSet(CacheGCFlag.Name) {
-		cache.TrieNodeLimit = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheGCFlag.Name) / 100
+		cache.TrieDirtyLimit = ctx.GlobalInt(CacheFlag.Name) * ctx.GlobalInt(CacheGCFlag.Name) / 100
 	}
 	vmcfg := vm.Config{EnablePreimageRecording: ctx.GlobalBool(VMEnableDebugFlag.Name)}
 	chain, err = core.NewBlockChain(chainDb, cache, config, engine, vmcfg, nil)

core/blockchain.go

@@ -68,9 +68,10 @@ const (
 // CacheConfig contains the configuration values for the trie caching/pruning
 // that's resident in a blockchain.
 type CacheConfig struct {
-	Disabled      bool          // Whether to disable trie write caching (archive node)
-	TrieNodeLimit int           // Memory limit (MB) at which to flush the current in-memory trie to disk
-	TrieTimeLimit time.Duration // Time limit after which to flush the current in-memory trie to disk
+	Disabled       bool          // Whether to disable trie write caching (archive node)
+	TrieCleanLimit int           // Memory allowance (MB) to use for caching trie nodes in memory
+	TrieDirtyLimit int           // Memory limit (MB) at which to start flushing dirty trie nodes to disk
+	TrieTimeLimit  time.Duration // Time limit after which to flush the current in-memory trie to disk
 }
 
 // BlockChain represents the canonical chain given a database with a genesis
@@ -140,8 +141,9 @@ type BlockChain struct {
 func NewBlockChain(db ethdb.Database, cacheConfig *CacheConfig, chainConfig *params.ChainConfig, engine consensus.Engine, vmConfig vm.Config, shouldPreserve func(block *types.Block) bool) (*BlockChain, error) {
 	if cacheConfig == nil {
 		cacheConfig = &CacheConfig{
-			TrieNodeLimit: 256,
-			TrieTimeLimit: 5 * time.Minute,
+			TrieCleanLimit: 256,
+			TrieDirtyLimit: 256,
+			TrieTimeLimit:  5 * time.Minute,
 		}
 	}
 	bodyCache, _ := lru.New(bodyCacheLimit)
@@ -156,7 +158,7 @@ func NewBlockChain(db ethdb.Database, cacheConfig *CacheConfig, chainConfig *par
 		cacheConfig:    cacheConfig,
 		db:             db,
 		triegc:         prque.New(nil),
-		stateCache:     state.NewDatabase(db),
+		stateCache:     state.NewDatabaseWithCache(db, cacheConfig.TrieCleanLimit),
 		quit:           make(chan struct{}),
 		shouldPreserve: shouldPreserve,
 		bodyCache:      bodyCache,
@@ -393,6 +395,11 @@ func (bc *BlockChain) StateAt(root common.Hash) (*state.StateDB, error) {
 	return state.New(root, bc.stateCache)
 }
 
+// StateCache returns the caching database underpinning the blockchain instance.
+func (bc *BlockChain) StateCache() state.Database {
+	return bc.stateCache
+}
+
 // Reset purges the entire blockchain, restoring it to its genesis state.
 func (bc *BlockChain) Reset() error {
 	return bc.ResetWithGenesisBlock(bc.genesisBlock)
@@ -938,7 +945,7 @@ func (bc *BlockChain) WriteBlockWithState(block *types.Block, receipts []*types.
 			// If we exceeded our memory allowance, flush matured singleton nodes to disk
 			var (
 				nodes, imgs = triedb.Size()
-				limit       = common.StorageSize(bc.cacheConfig.TrieNodeLimit) * 1024 * 1024
+				limit       = common.StorageSize(bc.cacheConfig.TrieDirtyLimit) * 1024 * 1024
 			)
 			if nodes > limit || imgs > 4*1024*1024 {
 				triedb.Cap(limit - ethdb.IdealBatchSize)

core/state/database.go

@@ -72,13 +72,19 @@ type Trie interface {
 }
 
 // NewDatabase creates a backing store for state. The returned database is safe for
-// concurrent use and retains cached trie nodes in memory. The pool is an optional
-// intermediate trie-node memory pool between the low level storage layer and the
-// high level trie abstraction.
+// concurrent use and retains a few recent expanded trie nodes in memory. To keep
+// more historical state in memory, use the NewDatabaseWithCache constructor.
 func NewDatabase(db ethdb.Database) Database {
+	return NewDatabaseWithCache(db, 0)
+}
+
+// NewDatabase creates a backing store for state. The returned database is safe for
+// concurrent use and retains both a few recent expanded trie nodes in memory, as
+// well as a lot of collapsed RLP trie nodes in a large memory cache.
+func NewDatabaseWithCache(db ethdb.Database, cache int) Database {
 	csc, _ := lru.New(codeSizeCacheSize)
 	return &cachingDB{
-		db:            trie.NewDatabase(db),
+		db:            trie.NewDatabaseWithCache(db, cache),
 		codeSizeCache: csc,
 	}
 }

eth/api.go

@@ -444,16 +444,16 @@ func (api *PrivateDebugAPI) getModifiedAccounts(startBlock, endBlock *types.Bloc
 	if startBlock.Number().Uint64() >= endBlock.Number().Uint64() {
 		return nil, fmt.Errorf("start block height (%d) must be less than end block height (%d)", startBlock.Number().Uint64(), endBlock.Number().Uint64())
 	}
+	triedb := api.eth.BlockChain().StateCache().TrieDB()
 
-	oldTrie, err := trie.NewSecure(startBlock.Root(), trie.NewDatabase(api.eth.chainDb), 0)
+	oldTrie, err := trie.NewSecure(startBlock.Root(), triedb, 0)
 	if err != nil {
 		return nil, err
 	}
-	newTrie, err := trie.NewSecure(endBlock.Root(), trie.NewDatabase(api.eth.chainDb), 0)
+	newTrie, err := trie.NewSecure(endBlock.Root(), triedb, 0)
 	if err != nil {
 		return nil, err
 	}
-
 	diff, _ := trie.NewDifferenceIterator(oldTrie.NodeIterator([]byte{}), newTrie.NodeIterator([]byte{}))
 	iter := trie.NewIterator(diff)
 

eth/api_tracer.go

@@ -138,7 +138,7 @@ func (api *PrivateDebugAPI) traceChain(ctx context.Context, start, end *types.Bl
 
 	// Ensure we have a valid starting state before doing any work
 	origin := start.NumberU64()
-	database := state.NewDatabase(api.eth.ChainDb())
+	database := state.NewDatabaseWithCache(api.eth.ChainDb(), 16) // Chain tracing will probably start at genesis
 
 	if number := start.NumberU64(); number > 0 {
 		start = api.eth.blockchain.GetBlock(start.ParentHash(), start.NumberU64()-1)
@@ -492,7 +492,7 @@ func (api *PrivateDebugAPI) computeStateDB(block *types.Block, reexec uint64) (*
 	}
 	// Otherwise try to reexec blocks until we find a state or reach our limit
 	origin := block.NumberU64()
-	database := state.NewDatabase(api.eth.ChainDb())
+	database := state.NewDatabaseWithCache(api.eth.ChainDb(), 16)
 
 	for i := uint64(0); i < reexec; i++ {
 		block = api.eth.blockchain.GetBlock(block.ParentHash(), block.NumberU64()-1)

eth/backend.go

@@ -154,7 +154,7 @@ func New(ctx *node.ServiceContext, config *Config) (*Ethereum, error) {
 			EWASMInterpreter:        config.EWASMInterpreter,
 			EVMInterpreter:          config.EVMInterpreter,
 		}
-		cacheConfig = &core.CacheConfig{Disabled: config.NoPruning, TrieNodeLimit: config.TrieCache, TrieTimeLimit: config.TrieTimeout}
+		cacheConfig = &core.CacheConfig{Disabled: config.NoPruning, TrieCleanLimit: config.TrieCleanCache, TrieDirtyLimit: config.TrieDirtyCache, TrieTimeLimit: config.TrieTimeout}
 	)
 	eth.blockchain, err = core.NewBlockChain(chainDb, cacheConfig, eth.chainConfig, eth.engine, vmConfig, eth.shouldPreserve)
 	if err != nil {

eth/config.go

@@ -43,15 +43,16 @@ var DefaultConfig = Config{
 		DatasetsInMem:  1,
 		DatasetsOnDisk: 2,
 	},
-	NetworkId:     1,
-	LightPeers:    100,
-	DatabaseCache: 768,
-	TrieCache:     256,
-	TrieTimeout:   60 * time.Minute,
-	MinerGasFloor: 8000000,
-	MinerGasCeil:  8000000,
-	MinerGasPrice: big.NewInt(params.GWei),
-	MinerRecommit: 3 * time.Second,
+	NetworkId:      1,
+	LightPeers:     100,
+	DatabaseCache:  512,
+	TrieCleanCache: 256,
+	TrieDirtyCache: 256,
+	TrieTimeout:    60 * time.Minute,
+	MinerGasFloor:  8000000,
+	MinerGasCeil:   8000000,
+	MinerGasPrice:  big.NewInt(params.GWei),
+	MinerRecommit:  3 * time.Second,
 
 	TxPool: core.DefaultTxPoolConfig,
 	GPO: gasprice.Config{
@@ -94,7 +95,8 @@ type Config struct {
 	SkipBcVersionCheck bool `toml:"-"`
 	DatabaseHandles    int  `toml:"-"`
 	DatabaseCache      int
-	TrieCache          int
+	TrieCleanCache     int
+	TrieDirtyCache     int
 	TrieTimeout        time.Duration
 
 	// Mining-related options

light/postprocess.go

@@ -159,7 +159,7 @@ func NewChtIndexer(db ethdb.Database, odr OdrBackend, size, confirms uint64) *co
 		diskdb:      db,
 		odr:         odr,
 		trieTable:   trieTable,
-		triedb:      trie.NewDatabase(trieTable),
+		triedb:      trie.NewDatabaseWithCache(trieTable, 1), // Use a tiny cache only to keep memory down
 		sectionSize: size,
 	}
 	return core.NewChainIndexer(db, ethdb.NewTable(db, "chtIndex-"), backend, size, confirms, time.Millisecond*100, "cht")
@@ -281,7 +281,7 @@ func NewBloomTrieIndexer(db ethdb.Database, odr OdrBackend, parentSize, size uin
 		diskdb:     db,
 		odr:        odr,
 		trieTable:  trieTable,
-		triedb:     trie.NewDatabase(trieTable),
+		triedb:     trie.NewDatabaseWithCache(trieTable, 1), // Use a tiny cache only to keep memory down
 		parentSize: parentSize,
 		size:       size,
 	}

tests/block_test_util.go

@@ -118,7 +118,7 @@ func (t *BlockTest) Run() error {
 	} else {
 		engine = ethash.NewShared()
 	}
-	chain, err := core.NewBlockChain(db, nil, config, engine, vm.Config{}, nil)
+	chain, err := core.NewBlockChain(db, &core.CacheConfig{TrieCleanLimit: 0}, config, engine, vm.Config{}, nil)
 	if err != nil {
 		return err
 	}