Erigon 3: Prune canonical markers - CanonicalHash, HeaderNumber f…

…rom chaindata (#11634)

We prune `CanonicalHash` and `HeaderNumber ` in `chaindata`. this leads
to a -99% in these 2 tables and reduce their total size to 40 pages in
TOTAL (from >500k pages).

Benchmarks:

```
Total time taken to query blocks from 1000000 to 3000000: 1720.1410 seconds (OLD)
Total time taken to query blocks from 1000000 to 3000000: 1747.4729 seconds (New)

Amortized delta cost: 0.1ms/req
```

Is the RPC working for those ranges? Yes.

## How does it work?

Query canonical hash from snapshots by: `viewHeaderByNumber` ->
`header.Hash()` if db hit is missed
Query header number from snapshots by: `viewHeaderByHash` ->
`header.Num64` if db hit is missed

cmd/rpcdaemon/rpcservices/eth_backend.go

@@ -298,6 +298,12 @@ func (back *RemoteBackend) HeaderByHash(ctx context.Context, tx kv.Getter, hash
 func (back *RemoteBackend) CanonicalHash(ctx context.Context, tx kv.Getter, blockNum uint64) (common.Hash, error) {
 	return back.blockReader.CanonicalHash(ctx, tx, blockNum)
 }
+func (back *RemoteBackend) HeaderNumber(ctx context.Context, tx kv.Getter, hash common.Hash) (*uint64, error) {
+	return back.blockReader.HeaderNumber(ctx, tx, hash)
+}
+func (back *RemoteBackend) IsCanonical(ctx context.Context, tx kv.Getter, hash common.Hash, blockNum uint64) (bool, error) {
+	return back.blockReader.IsCanonical(ctx, tx, hash, blockNum)
+}
 func (back *RemoteBackend) TxnByIdxInBlock(ctx context.Context, tx kv.Getter, blockNum uint64, i int) (types.Transaction, error) {
 	return back.blockReader.TxnByIdxInBlock(ctx, tx, blockNum, i)
 }

core/rawdb/accessors_chain.go

@@ -483,20 +483,6 @@ func WriteBodyForStorage(db kv.Putter, hash common.Hash, number uint64, body *ty
 	return db.Put(kv.BlockBody, dbutils.BlockBodyKey(number, hash), data)
 }
 
-// ReadBodyByNumber - returns canonical block body
-func ReadBodyByNumber(db kv.Tx, number uint64) (*types.Body, uint64, uint32, error) {
-	hash, err := ReadCanonicalHash(db, number)
-	if err != nil {
-		return nil, 0, 0, fmt.Errorf("failed ReadCanonicalHash: %w", err)
-	}
-	if hash == (common.Hash{}) {
-		return nil, 0, 0, nil
-	}
-	body, baseTxnID, txCount := ReadBody(db, hash, number)
-	// TODO baseTxnID from ReadBody is baseTxnID+1, but we could expect original baseTxnID here
-	return body, baseTxnID, txCount, nil
-}
-
 func ReadBodyWithTransactions(db kv.Getter, hash common.Hash, number uint64) (*types.Body, error) {
 	body, firstTxId, txCount := ReadBody(db, hash, number)
 	if body == nil {

erigon-lib/common/dbg/experiments.go

@@ -45,10 +45,11 @@ var (
 	mergeTr = EnvInt("MERGE_THRESHOLD", -1)
 
 	//state v3
-	noPrune           = EnvBool("NO_PRUNE", false)
-	noMerge           = EnvBool("NO_MERGE", false)
-	discardHistory    = EnvBool("DISCARD_HISTORY", false)
-	discardCommitment = EnvBool("DISCARD_COMMITMENT", false)
+	noPrune              = EnvBool("NO_PRUNE", false)
+	noMerge              = EnvBool("NO_MERGE", false)
+	discardHistory       = EnvBool("DISCARD_HISTORY", false)
+	discardCommitment    = EnvBool("DISCARD_COMMITMENT", false)
+	pruneTotalDifficulty = EnvBool("PRUNE_TOTAL_DIFFICULTY", false)
 
 	// force skipping of any non-Erigon2 .torrent files
 	DownloaderOnlyBlocks = EnvBool("DOWNLOADER_ONLY_BLOCKS", false)
@@ -82,10 +83,11 @@ func NoSync() bool        { return noSync }
 func MdbxReadAhead() bool { return mdbxReadahead }
 func MdbxLockInRam() bool { return mdbxLockInRam }
 
-func DiscardHistory() bool    { return discardHistory }
-func DiscardCommitment() bool { return discardCommitment }
-func NoPrune() bool           { return noPrune }
-func NoMerge() bool           { return noMerge }
+func DiscardHistory() bool       { return discardHistory }
+func DiscardCommitment() bool    { return discardCommitment }
+func NoPrune() bool              { return noPrune }
+func NoMerge() bool              { return noMerge }
+func PruneTotalDifficulty() bool { return pruneTotalDifficulty }
 
 var (
 	dirtySace     uint64

erigon-lib/direct/eth_backend_client.go

@@ -191,6 +191,14 @@ func (c *SubscribeLogsStreamC) Recv() (*remote.SubscribeLogsReply, error) {
 
 // -- end SubscribeLogs
 
+func (s *EthBackendClientDirect) CanonicalHash(ctx context.Context, in *remote.CanonicalHashRequest, opts ...grpc.CallOption) (*remote.CanonicalHashReply, error) {
+	return s.server.CanonicalHash(ctx, in)
+}
+
+func (s *EthBackendClientDirect) HeaderNumber(ctx context.Context, in *remote.HeaderNumberRequest, opts ...grpc.CallOption) (*remote.HeaderNumberReply, error) {
+	return s.server.HeaderNumber(ctx, in)
+}
+
 func (s *EthBackendClientDirect) Block(ctx context.Context, in *remote.BlockRequest, opts ...grpc.CallOption) (*remote.BlockReply, error) {
 	return s.server.Block(ctx, in)
 }

erigon-lib/go.mod

@@ -4,7 +4,7 @@ go 1.21.5
 
 require (
 	github.com/erigontech/erigon-snapshot v1.3.1-0.20240814160410-2ce37904b978
-	github.com/erigontech/interfaces v0.0.0-20240723225543-c6b574058f8c
+	github.com/erigontech/interfaces v0.0.0-20240816222220-760b11456e56
 	github.com/erigontech/mdbx-go v0.38.4
 	github.com/erigontech/secp256k1 v1.1.0
 	github.com/rs/dnscache v0.0.0-20211102005908-e0241e321417

erigon-lib/go.sum

@@ -148,8 +148,8 @@ github.com/envoyproxy/go-control-plane v0.9.4/go.mod h1:6rpuAdCZL397s3pYoYcLgu1m
 github.com/envoyproxy/protoc-gen-validate v0.1.0/go.mod h1:iSmxcyjqTsJpI2R4NaDN7+kN2VEUnK/pcBlmesArF7c=
 github.com/erigontech/erigon-snapshot v1.3.1-0.20240814160410-2ce37904b978 h1:7ECOf7Us3+/706WGZXIX84qQc6zmxQby8fGbFLiqFlU=
 github.com/erigontech/erigon-snapshot v1.3.1-0.20240814160410-2ce37904b978/go.mod h1:ooHlCl+eEYzebiPu+FP6Q6SpPUeMADn8Jxabv3IKb9M=
-github.com/erigontech/interfaces v0.0.0-20240723225543-c6b574058f8c h1:KbcMdRKYRL+A4bJGK+fNOITk95wJzh9rUXFJ17wCyUY=
-github.com/erigontech/interfaces v0.0.0-20240723225543-c6b574058f8c/go.mod h1:N7OUkhkcagp9+7yb4ycHsG2VWCOmuJ1ONBecJshxtLE=
+github.com/erigontech/interfaces v0.0.0-20240816222220-760b11456e56 h1:RFLd10LsbUdcXS7zntLuVj+mZZJFDD6E7y183kYC+SM=
+github.com/erigontech/interfaces v0.0.0-20240816222220-760b11456e56/go.mod h1:N7OUkhkcagp9+7yb4ycHsG2VWCOmuJ1ONBecJshxtLE=
 github.com/erigontech/mdbx-go v0.38.4 h1:S9T7mTe9KPcFe4dOoOtVdI6gPzht9y7wMnYfUBgrQLo=
 github.com/erigontech/mdbx-go v0.38.4/go.mod h1:IcOLQDPw3VM/asP6T5JVPPN4FHHgJtY16XfYjzWKVNI=
 github.com/erigontech/secp256k1 v1.1.0 h1:mO3YJMUSoASE15Ya//SoHiisptUhdXExuMUN1M0X9qY=