core/state: state size tracking (ethereum#32362)

Add state size tracking and retrieve api, start geth with `--state.size-tracking`, 
the initial bootstrap is required (around 1h on mainnet), after the bootstrap, 
use `debug_stateSize()` RPC to retrieve the state size:

```
> debug.stateSize()
{
  accountBytes: "0x39681967b",
  accountTrienodeBytes: "0xc57939f0c",
  accountTrienodes: "0x198b36ac",
  accounts: "0x129da14a",
  blockNumber: "0x1635e90",
  contractCodeBytes: "0x2b63ef481",
  contractCodes: "0x1c7b45",
  stateRoot: "0x9c36a3ec3745d72eea8700bd27b90dcaa66de0494b187c5600750044151e620a",
  storageBytes: "0x18a6e7d3f1",
  storageTrienodeBytes: "0x2e7f53fae6",
  storageTrienodes: "0x6e49a234",
  storages: "0x517859c5"
}
```

---------

Signed-off-by: jsvisa <[email protected]>
Co-authored-by: Gary Rong <[email protected]>

Author: jsvisa

cmd/geth/chaincmd.go

@@ -108,6 +108,7 @@ if one is set.  Otherwise it prints the genesis from the datadir.`,
 			utils.MetricsInfluxDBTokenFlag,
 			utils.MetricsInfluxDBBucketFlag,
 			utils.MetricsInfluxDBOrganizationFlag,
+			utils.StateSizeTrackingFlag,
 			utils.TxLookupLimitFlag,
 			utils.VMTraceFlag,
 			utils.VMTraceJsonConfigFlag,

cmd/geth/config.go

@@ -209,7 +209,7 @@ func constructDevModeBanner(ctx *cli.Context, cfg gethConfig) string {
        0x%x (10^49 ETH)
 `, cfg.Eth.Miner.PendingFeeRecipient)
 		if cfg.Eth.Miner.PendingFeeRecipient == utils.DeveloperAddr {
-			devModeBanner += fmt.Sprintf(` 
+			devModeBanner += fmt.Sprintf(`
        Private Key
        ------------------
        0x%x

cmd/geth/main.go

@@ -200,6 +200,7 @@ var (
 		utils.MetricsInfluxDBTokenFlag,
 		utils.MetricsInfluxDBBucketFlag,
 		utils.MetricsInfluxDBOrganizationFlag,
+		utils.StateSizeTrackingFlag,
 	}
 )
 

cmd/utils/flags.go

@@ -270,6 +270,12 @@ var (
 		Usage:    "Scheme to use for storing ethereum state ('hash' or 'path')",
 		Category: flags.StateCategory,
 	}
+	StateSizeTrackingFlag = &cli.BoolFlag{
+		Name:     "state.size-tracking",
+		Usage:    "Enable state size tracking, retrieve state size with debug_stateSize.",
+		Value:    ethconfig.Defaults.EnableStateSizeTracking,
+		Category: flags.StateCategory,
+	}
 	StateHistoryFlag = &cli.Uint64Flag{
 		Name:     "history.state",
 		Usage:    "Number of recent blocks to retain state history for, only relevant in state.scheme=path (default = 90,000 blocks, 0 = entire chain)",
@@ -1726,6 +1732,9 @@ func SetEthConfig(ctx *cli.Context, stack *node.Node, cfg *ethconfig.Config) {
 			cfg.EthDiscoveryURLs = SplitAndTrim(urls)
 		}
 	}
+	if ctx.Bool(StateSizeTrackingFlag.Name) {
+		cfg.EnableStateSizeTracking = true
+	}
 	// Override any default configs for hard coded networks.
 	switch {
 	case ctx.Bool(MainnetFlag.Name):
@@ -2208,6 +2217,9 @@ func MakeChain(ctx *cli.Context, stack *node.Node, readonly bool) (*core.BlockCh
 		// - DATADIR/triedb/merkle.journal
 		// - DATADIR/triedb/verkle.journal
 		TrieJournalDirectory: stack.ResolvePath("triedb"),
+
+		// Enable state size tracking if enabled
+		StateSizeTracking: ctx.Bool(StateSizeTrackingFlag.Name),
 	}
 	if options.ArchiveMode && !options.Preimages {
 		options.Preimages = true

core/blockchain.go

@@ -196,6 +196,9 @@ type BlockChainConfig struct {
 	// If the value is zero, all transactions of the entire chain will be indexed.
 	// If the value is -1, indexing is disabled.
 	TxLookupLimit int64
+
+	// StateSizeTracking indicates whether the state size tracking is enabled.
+	StateSizeTracking bool
 }
 
 // DefaultConfig returns the default config.
@@ -333,6 +336,7 @@ type BlockChain struct {
 	prefetcher Prefetcher
 	processor  Processor // Block transaction processor interface
 	logger     *tracing.Hooks
+	stateSizer *state.SizeTracker // State size tracking
 
 	lastForkReadyAlert time.Time // Last time there was a fork readiness print out
 }
@@ -526,6 +530,17 @@ func NewBlockChain(db ethdb.Database, genesis *Genesis, engine consensus.Engine,
 	if bc.cfg.TxLookupLimit >= 0 {
 		bc.txIndexer = newTxIndexer(uint64(bc.cfg.TxLookupLimit), bc)
 	}
+
+	// Start state size tracker
+	if bc.cfg.StateSizeTracking {
+		stateSizer, err := state.NewSizeTracker(bc.db, bc.triedb)
+		if err == nil {
+			bc.stateSizer = stateSizer
+			log.Info("Enabled state size metrics")
+		} else {
+			log.Info("Failed to setup size tracker", "err", err)
+		}
+	}
 	return bc, nil
 }
 
@@ -1252,6 +1267,10 @@ func (bc *BlockChain) stopWithoutSaving() {
 	// Signal shutdown to all goroutines.
 	bc.InterruptInsert(true)
 
+	// Stop state size tracker
+	if bc.stateSizer != nil {
+		bc.stateSizer.Stop()
+	}
 	// Now wait for all chain modifications to end and persistent goroutines to exit.
 	//
 	// Note: Close waits for the mutex to become available, i.e. any running chain
@@ -1586,10 +1605,14 @@ func (bc *BlockChain) writeBlockWithState(block *types.Block, receipts []*types.
 		log.Crit("Failed to write block into disk", "err", err)
 	}
 	// Commit all cached state changes into underlying memory database.
-	root, err := statedb.Commit(block.NumberU64(), bc.chainConfig.IsEIP158(block.Number()), bc.chainConfig.IsCancun(block.Number(), block.Time()))
+	root, stateUpdate, err := statedb.CommitWithUpdate(block.NumberU64(), bc.chainConfig.IsEIP158(block.Number()), bc.chainConfig.IsCancun(block.Number(), block.Time()))
 	if err != nil {
 		return err
 	}
+	// Emit the state update to the state sizestats if it's active
+	if bc.stateSizer != nil {
+		bc.stateSizer.Notify(stateUpdate)
+	}
 	// If node is running in path mode, skip explicit gc operation
 	// which is unnecessary in this mode.
 	if bc.triedb.Scheme() == rawdb.PathScheme {
@@ -2791,3 +2814,8 @@ func (bc *BlockChain) SetTrieFlushInterval(interval time.Duration) {
 func (bc *BlockChain) GetTrieFlushInterval() time.Duration {
 	return time.Duration(bc.flushInterval.Load())
 }
+
+// StateSizer returns the state size tracker, or nil if it's not initialized
+func (bc *BlockChain) StateSizer() *state.SizeTracker {
+	return bc.stateSizer
+}