core: reduce load on txindexer from API

core/blockchain_reader.go

@@ -270,42 +270,20 @@ func (bc *BlockChain) GetAncestor(hash common.Hash, number, ancestor uint64, max
 // GetTransactionLookup retrieves the lookup along with the transaction
 // itself associate with the given transaction hash.
 //
-// An error will be returned if the transaction is not found, and background
-// indexing for transactions is still in progress. The transaction might be
-// reachable shortly once it's indexed.
-//
-// A null will be returned in the transaction is not found and background
-// transaction indexing is already finished. The transaction is not existent
-// from the node's perspective.
-func (bc *BlockChain) GetTransactionLookup(hash common.Hash) (*rawdb.LegacyTxLookupEntry, *types.Transaction, error) {
+// A null will be returned if the transaction is not found. This can be due to
+// the transaction indexer not being finished. The caller must explicitly check
+// the indexer progress.
+func (bc *BlockChain) GetTransactionLookup(hash common.Hash) (*rawdb.LegacyTxLookupEntry, *types.Transaction) {
 	bc.txLookupLock.RLock()
 	defer bc.txLookupLock.RUnlock()
 
 	// Short circuit if the txlookup already in the cache, retrieve otherwise
 	if item, exist := bc.txLookupCache.Get(hash); exist {
-		return item.lookup, item.transaction, nil
+		return item.lookup, item.transaction
 	}
 	tx, blockHash, blockNumber, txIndex := rawdb.ReadTransaction(bc.db, hash)
 	if tx == nil {
-		progress, err := bc.TxIndexProgress()
-		if err != nil {
-			// No error is returned if the transaction indexing progress is unreachable
-			// due to unexpected internal errors. In such cases, it is impossible to
-			// determine whether the transaction does not exist or has simply not been
-			// indexed yet without a progress marker.
-			//
-			// In such scenarios, the transaction is treated as unreachable, though
-			// this is clearly an unintended and unexpected situation.
-			return nil, nil, nil
-		}
-		// The transaction indexing is not finished yet, returning an
-		// error to explicitly indicate it.
-		if !progress.Done() {
-			return nil, nil, errors.New("transaction indexing still in progress")
-		}
-		// The transaction is already indexed, the transaction is either
-		// not existent or not in the range of index, returning null.
-		return nil, nil, nil
+		return nil, nil
 	}
 	lookup := &rawdb.LegacyTxLookupEntry{
 		BlockHash:  blockHash,
@@ -316,7 +294,23 @@ func (bc *BlockChain) GetTransactionLookup(hash common.Hash) (*rawdb.LegacyTxLoo
 		lookup:      lookup,
 		transaction: tx,
 	})
-	return lookup, tx, nil
+	return lookup, tx
+}
+
+// TxIndexDone returns true if the transaction indexer has finished indexing.
+func (bc *BlockChain) TxIndexDone() bool {
+	progress, err := bc.TxIndexProgress()
+	if err != nil {
+		// No error is returned if the transaction indexing progress is unreachable
+		// due to unexpected internal errors. In such cases, it is impossible to
+		// determine whether the transaction does not exist or has simply not been
+		// indexed yet without a progress marker.
+		//
+		// In such scenarios, the transaction is treated as unreachable, though
+		// this is clearly an unintended and unexpected situation.
+		return true
+	}
+	return progress.Done()
 }
 
 // HasState checks if state trie is fully present in the database or not.
@@ -412,7 +406,7 @@ func (bc *BlockChain) TxIndexProgress() (TxIndexProgress, error) {
 	if bc.txIndexer == nil {
 		return TxIndexProgress{}, errors.New("tx indexer is not enabled")
 	}
-	return bc.txIndexer.txIndexProgress()
+	return bc.txIndexer.txIndexProgress(), nil
 }
 
 // HistoryPruningCutoff returns the configured history pruning point.

core/txindexer.go

@@ -17,8 +17,8 @@
 package core
 
 import (
-	"errors"
 	"fmt"
+	"sync/atomic"
 
 	"github.com/ethereum/go-ethereum/common"
 	"github.com/ethereum/go-ethereum/core/rawdb"
@@ -47,26 +47,35 @@ type txIndexer struct {
 	//       and all others shouldn't.
 	limit uint64
 
+	// The current tail of the indexed transactions, null indicates
+	// that no transactions have been indexed yet.
+	//
+	// This field is accessed by both the indexer and the indexing
+	// progress queries.
+	tail atomic.Pointer[uint64]
+
 	// cutoff denotes the block number before which the chain segment should
 	// be pruned and not available locally.
-	cutoff   uint64
-	db       ethdb.Database
-	progress chan chan TxIndexProgress
-	term     chan chan struct{}
-	closed   chan struct{}
+	cutoff uint64
+	chain  *BlockChain
+	db     ethdb.Database
+	term   chan chan struct{}
+	closed chan struct{}
 }
 
 // newTxIndexer initializes the transaction indexer.
 func newTxIndexer(limit uint64, chain *BlockChain) *txIndexer {
 	cutoff, _ := chain.HistoryPruningCutoff()
 	indexer := &txIndexer{
-		limit:    limit,
-		cutoff:   cutoff,
-		db:       chain.db,
-		progress: make(chan chan TxIndexProgress),
-		term:     make(chan chan struct{}),
-		closed:   make(chan struct{}),
+		limit:  limit,
+		cutoff: cutoff,
+		chain:  chain,
+		db:     chain.db,
+		term:   make(chan chan struct{}),
+		closed: make(chan struct{}),
 	}
+	indexer.tail.Store(rawdb.ReadTxIndexTail(chain.db))
+
 	go indexer.loop(chain)
 
 	var msg string
@@ -154,6 +163,7 @@ func (indexer *txIndexer) repair(head uint64) {
 		// A crash may occur between the two delete operations,
 		// potentially leaving dangling indexes in the database.
 		// However, this is considered acceptable.
+		indexer.tail.Store(nil)
 		rawdb.DeleteTxIndexTail(indexer.db)
 		rawdb.DeleteAllTxLookupEntries(indexer.db, nil)
 		log.Warn("Purge transaction indexes", "head", head, "tail", *tail)
@@ -174,6 +184,7 @@ func (indexer *txIndexer) repair(head uint64) {
 		// Traversing the database directly within the transaction
 		// index namespace might be slow and expensive, but we
 		// have no choice.
+		indexer.tail.Store(nil)
 		rawdb.DeleteTxIndexTail(indexer.db)
 		rawdb.DeleteAllTxLookupEntries(indexer.db, nil)
 		log.Warn("Purge transaction indexes", "head", head, "cutoff", indexer.cutoff)
@@ -187,6 +198,7 @@ func (indexer *txIndexer) repair(head uint64) {
 		// A crash may occur between the two delete operations,
 		// potentially leaving dangling indexes in the database.
 		// However, this is considered acceptable.
+		indexer.tail.Store(&indexer.cutoff)
 		rawdb.WriteTxIndexTail(indexer.db, indexer.cutoff)
 		rawdb.DeleteAllTxLookupEntries(indexer.db, func(txhash common.Hash, blob []byte) bool {
 			n := rawdb.DecodeTxLookupEntry(blob, indexer.db)
@@ -243,12 +255,10 @@ func (indexer *txIndexer) loop(chain *BlockChain) {
 				done = make(chan struct{})
 				go indexer.run(h.Header.Number.Uint64(), stop, done)
 			}
-			head = h.Header.Number.Uint64()
 		case <-done:
 			stop = nil
 			done = nil
-		case ch := <-indexer.progress:
-			ch <- indexer.report(head)
+			indexer.tail.Store(rawdb.ReadTxIndexTail(indexer.db))
 		case ch := <-indexer.term:
 			if stop != nil {
 				close(stop)
@@ -264,7 +274,7 @@ func (indexer *txIndexer) loop(chain *BlockChain) {
 }
 
 // report returns the tx indexing progress.
-func (indexer *txIndexer) report(head uint64) TxIndexProgress {
+func (indexer *txIndexer) report(head uint64, tail *uint64) TxIndexProgress {
 	// Special case if the head is even below the cutoff,
 	// nothing to index.
 	if head < indexer.cutoff {
@@ -284,7 +294,6 @@ func (indexer *txIndexer) report(head uint64) TxIndexProgress {
 	}
 	// Compute how many blocks have been indexed
 	var indexed uint64
-	tail := rawdb.ReadTxIndexTail(indexer.db)
 	if tail != nil {
 		indexed = head - *tail + 1
 	}
@@ -300,16 +309,12 @@ func (indexer *txIndexer) report(head uint64) TxIndexProgress {
 	}
 }
 
-// txIndexProgress retrieves the tx indexing progress, or an error if the
-// background tx indexer is already stopped.
-func (indexer *txIndexer) txIndexProgress() (TxIndexProgress, error) {
-	ch := make(chan TxIndexProgress, 1)
-	select {
-	case indexer.progress <- ch:
-		return <-ch, nil
-	case <-indexer.closed:
-		return TxIndexProgress{}, errors.New("indexer is closed")
-	}
+// txIndexProgress retrieves the transaction indexing progress. The reported
+// progress may slightly lag behind the actual indexing state, as the tail is
+// only updated at the end of each indexing operation. However, this delay is
+// considered acceptable.
+func (indexer *txIndexer) txIndexProgress() TxIndexProgress {
+	return indexer.report(indexer.resolveHead(), indexer.tail.Load())
 }
 
 // close shutdown the indexer. Safe to be called for multiple times.

core/txindexer_test.go

@@ -121,9 +121,8 @@ func TestTxIndexer(t *testing.T) {
 
 		// Index the initial blocks from ancient store
 		indexer := &txIndexer{
-			limit:    0,
-			db:       db,
-			progress: make(chan chan TxIndexProgress),
+			limit: 0,
+			db:    db,
 		}
 		for i, limit := range c.limits {
 			indexer.limit = limit
@@ -241,9 +240,8 @@ func TestTxIndexerRepair(t *testing.T) {
 
 		// Index the initial blocks from ancient store
 		indexer := &txIndexer{
-			limit:    c.limit,
-			db:       db,
-			progress: make(chan chan TxIndexProgress),
+			limit: c.limit,
+			db:    db,
 		}
 		indexer.run(chainHead, make(chan struct{}), make(chan struct{}))
 
@@ -432,15 +430,11 @@ func TestTxIndexerReport(t *testing.T) {
 
 		// Index the initial blocks from ancient store
 		indexer := &txIndexer{
-			limit:    c.limit,
-			cutoff:   c.cutoff,
-			db:       db,
-			progress: make(chan chan TxIndexProgress),
+			limit:  c.limit,
+			cutoff: c.cutoff,
+			db:     db,
 		}
-		if c.tail != nil {
-			rawdb.WriteTxIndexTail(db, *c.tail)
-		}
-		p := indexer.report(c.head)
+		p := indexer.report(c.head, c.tail)
 		if p.Indexed != c.expIndexed {
 			t.Fatalf("Unexpected indexed: %d, expected: %d", p.Indexed, c.expIndexed)
 		}

eth/api_backend.go

@@ -349,22 +349,20 @@ func (b *EthAPIBackend) GetPoolTransaction(hash common.Hash) *types.Transaction
 // GetTransaction retrieves the lookup along with the transaction itself associate
 // with the given transaction hash.
 //
-// An error will be returned if the transaction is not found, and background
-// indexing for transactions is still in progress. The error is used to indicate the
-// scenario explicitly that the transaction might be reachable shortly.
-//
-// A null will be returned in the transaction is not found and background transaction
-// indexing is already finished. The transaction is not existent from the perspective
-// of node.
-func (b *EthAPIBackend) GetTransaction(ctx context.Context, txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64, error) {
-	lookup, tx, err := b.eth.blockchain.GetTransactionLookup(txHash)
-	if err != nil {
-		return false, nil, common.Hash{}, 0, 0, err
-	}
+// A null will be returned if the transaction is not found. The transaction is not
+// existent from the node's perspective. This can be due to the transaction indexer
+// not being finished. The caller must explicitly check the indexer progress.
+func (b *EthAPIBackend) GetTransaction(txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64) {
+	lookup, tx := b.eth.blockchain.GetTransactionLookup(txHash)
 	if lookup == nil || tx == nil {
-		return false, nil, common.Hash{}, 0, 0, nil
+		return false, nil, common.Hash{}, 0, 0
 	}
-	return true, tx, lookup.BlockHash, lookup.BlockIndex, lookup.Index, nil
+	return true, tx, lookup.BlockHash, lookup.BlockIndex, lookup.Index
+}
+
+// TxIndexDone returns true if the transaction indexer has finished indexing.
+func (b *EthAPIBackend) TxIndexDone() bool {
+	return b.eth.blockchain.TxIndexDone()
 }
 
 func (b *EthAPIBackend) GetPoolNonce(ctx context.Context, addr common.Address) (uint64, error) {
@@ -391,7 +389,7 @@ func (b *EthAPIBackend) SubscribeNewTxsEvent(ch chan<- core.NewTxsEvent) event.S
 	return b.eth.txPool.SubscribeTransactions(ch, true)
 }
 
-func (b *EthAPIBackend) SyncProgress() ethereum.SyncProgress {
+func (b *EthAPIBackend) SyncProgress(ctx context.Context) ethereum.SyncProgress {
 	prog := b.eth.Downloader().Progress()
 	if txProg, err := b.eth.blockchain.TxIndexProgress(); err == nil {
 		prog.TxIndexFinishedBlocks = txProg.Indexed

eth/tracers/api.go

@@ -82,7 +82,8 @@ type Backend interface {
 	HeaderByNumber(ctx context.Context, number rpc.BlockNumber) (*types.Header, error)
 	BlockByHash(ctx context.Context, hash common.Hash) (*types.Block, error)
 	BlockByNumber(ctx context.Context, number rpc.BlockNumber) (*types.Block, error)
-	GetTransaction(ctx context.Context, txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64, error)
+	GetTransaction(txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64)
+	TxIndexDone() bool
 	RPCGasCap() uint64
 	ChainConfig() *params.ChainConfig
 	Engine() consensus.Engine
@@ -858,12 +859,13 @@ func containsTx(block *types.Block, hash common.Hash) bool {
 // TraceTransaction returns the structured logs created during the execution of EVM
 // and returns them as a JSON object.
 func (api *API) TraceTransaction(ctx context.Context, hash common.Hash, config *TraceConfig) (interface{}, error) {
-	found, _, blockHash, blockNumber, index, err := api.backend.GetTransaction(ctx, hash)
-	if err != nil {
-		return nil, ethapi.NewTxIndexingError()
-	}
-	// Only mined txes are supported
+	found, _, blockHash, blockNumber, index := api.backend.GetTransaction(hash)
 	if !found {
+		// Warn in case tx indexer is not done.
+		if !api.backend.TxIndexDone() {
+			return nil, ethapi.NewTxIndexingError()
+		}
+		// Only mined txes are supported
 		return nil, errTxNotFound
 	}
 	// It shouldn't happen in practice.

eth/tracers/api_test.go

@@ -116,9 +116,13 @@ func (b *testBackend) BlockByNumber(ctx context.Context, number rpc.BlockNumber)
 	return b.chain.GetBlockByNumber(uint64(number)), nil
 }
 
-func (b *testBackend) GetTransaction(ctx context.Context, txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64, error) {
+func (b *testBackend) GetTransaction(txHash common.Hash) (bool, *types.Transaction, common.Hash, uint64, uint64) {
 	tx, hash, blockNumber, index := rawdb.ReadTransaction(b.chaindb, txHash)
-	return tx != nil, tx, hash, blockNumber, index, nil
+	return tx != nil, tx, hash, blockNumber, index
+}
+
+func (b *testBackend) TxIndexDone() bool {
+	return true
 }
 
 func (b *testBackend) RPCGasCap() uint64 {

ethstats/ethstats.go

@@ -66,7 +66,7 @@ type backend interface {
 	CurrentHeader() *types.Header
 	HeaderByNumber(ctx context.Context, number rpc.BlockNumber) (*types.Header, error)
 	Stats() (pending int, queued int)
-	SyncProgress() ethereum.SyncProgress
+	SyncProgress(ctx context.Context) ethereum.SyncProgress
 }
 
 // fullNodeBackend encompasses the functionality necessary for a full node
@@ -766,7 +766,7 @@ func (s *Service) reportStats(conn *connWrapper) error {
 	)
 	// check if backend is a full node
 	if fullBackend, ok := s.backend.(fullNodeBackend); ok {
-		sync := fullBackend.SyncProgress()
+		sync := fullBackend.SyncProgress(context.Background())
 		syncing = !sync.Done()
 
 		price, _ := fullBackend.SuggestGasTipCap(context.Background())
@@ -775,7 +775,7 @@ func (s *Service) reportStats(conn *connWrapper) error {
 			gasprice += int(basefee.Uint64())
 		}
 	} else {
-		sync := s.backend.SyncProgress()
+		sync := s.backend.SyncProgress(context.Background())
 		syncing = !sync.Done()
 	}
 	// Assemble the node stats and send it to the server

graphql/graphql.go

@@ -229,7 +229,7 @@ func (t *Transaction) resolve(ctx context.Context) (*types.Transaction, *Block)
 		return t.tx, t.block
 	}
 	// Try to return an already finalized transaction
-	found, tx, blockHash, _, index, _ := t.r.backend.GetTransaction(ctx, t.hash)
+	found, tx, blockHash, _, index := t.r.backend.GetTransaction(t.hash)
 	if found {
 		t.tx = tx
 		blockNrOrHash := rpc.BlockNumberOrHashWithHash(blockHash, false)
@@ -1530,8 +1530,8 @@ func (s *SyncState) TxIndexRemainingBlocks() hexutil.Uint64 {
 // - healingBytecode:     number of bytecodes pending
 // - txIndexFinishedBlocks:  number of blocks whose transactions are indexed
 // - txIndexRemainingBlocks: number of blocks whose transactions are not indexed yet
-func (r *Resolver) Syncing() (*SyncState, error) {
-	progress := r.backend.SyncProgress()
+func (r *Resolver) Syncing(ctx context.Context) (*SyncState, error) {
+	progress := r.backend.SyncProgress(ctx)
 
 	// Return not syncing if the synchronisation already completed
 	if progress.Done() {