core, eth: improve delivery speed on header requests (ethereum#23105)

This PR reduces the amount of work we do when answering header queries, e.g. when a peer
is syncing from us.

For some items, e.g block bodies, when we read the rlp-data from database, we plug it
directly into the response package. We didn't do that for headers, but instead read
headers-rlp, decode to types.Header, and re-encode to rlp. This PR changes that to keep it
in RLP-form as much as possible. When a node is syncing from us, it typically requests 192
contiguous headers. On master it has the following effect:

- For headers not in ancient: 2 db lookups. One for translating hash->number (even though
  the request is by number), and another for reading by hash (this latter one is sometimes
  cached).
  
- For headers in ancient: 1 file lookup/syscall for translating hash->number (even though
  the request is by number), and another for reading the header itself. After this, it
  also performes a hashing of the header, to ensure that the hash is what it expected. In
  this PR, I instead move the logic for "give me a sequence of blocks" into the lower
  layers, where the database can determine how and what to read from leveldb and/or
  ancients.

There are basically four types of requests; three of them are improved this way. The
fourth, by hash going backwards, is more tricky to optimize. However, since we know that
the gap is 0, we can look up by the parentHash, and stlil shave off all the number->hash
lookups.

The gapped collection can be optimized similarly, as a follow-up, at least in three out of
four cases.

Co-authored-by: Felix Lange <fjl@twurst.com>

Author: 2 people

core/blockchain_reader.go

@@ -73,6 +73,12 @@ func (bc *BlockChain) GetHeaderByNumber(number uint64) *types.Header {
 	return bc.hc.GetHeaderByNumber(number)
 }
 
+// GetHeadersFrom returns a contiguous segment of headers, in rlp-form, going
+// backwards from the given number.
+func (bc *BlockChain) GetHeadersFrom(number, count uint64) []rlp.RawValue {
+	return bc.hc.GetHeadersFrom(number, count)
+}
+
 // GetBody retrieves a block body (transactions and uncles) from the database by
 // hash, caching it if found.
 func (bc *BlockChain) GetBody(hash common.Hash) *types.Body {

core/headerchain.go

@@ -33,6 +33,7 @@ import (
 	"github.com/ethereum/go-ethereum/ethdb"
 	"github.com/ethereum/go-ethereum/log"
 	"github.com/ethereum/go-ethereum/params"
+	"github.com/ethereum/go-ethereum/rlp"
 	lru "github.com/hashicorp/golang-lru"
 )
 
@@ -498,6 +499,46 @@ func (hc *HeaderChain) GetHeaderByNumber(number uint64) *types.Header {
 	return hc.GetHeader(hash, number)
 }
 
+// GetHeadersFrom returns a contiguous segment of headers, in rlp-form, going
+// backwards from the given number.
+// If the 'number' is higher than the highest local header, this method will
+// return a best-effort response, containing the headers that we do have.
+func (hc *HeaderChain) GetHeadersFrom(number, count uint64) []rlp.RawValue {
+	// If the request is for future headers, we still return the portion of
+	// headers that we are able to serve
+	if current := hc.CurrentHeader().Number.Uint64(); current < number {
+		if count > number-current {
+			count -= number - current
+			number = current
+		} else {
+			return nil
+		}
+	}
+	var headers []rlp.RawValue
+	// If we have some of the headers in cache already, use that before going to db.
+	hash := rawdb.ReadCanonicalHash(hc.chainDb, number)
+	if hash == (common.Hash{}) {
+		return nil
+	}
+	for count > 0 {
+		header, ok := hc.headerCache.Get(hash)
+		if !ok {
+			break
+		}
+		h := header.(*types.Header)
+		rlpData, _ := rlp.EncodeToBytes(h)
+		headers = append(headers, rlpData)
+		hash = h.ParentHash
+		count--
+		number--
+	}
+	// Read remaining from db
+	if count > 0 {
+		headers = append(headers, rawdb.ReadHeaderRange(hc.chainDb, number, count)...)
+	}
+	return headers
+}
+
 func (hc *HeaderChain) GetCanonicalHash(number uint64) common.Hash {
 	return rawdb.ReadCanonicalHash(hc.chainDb, number)
 }

core/rawdb/accessors_chain.go

@@ -279,6 +279,56 @@ func WriteFastTxLookupLimit(db ethdb.KeyValueWriter, number uint64) {
 	}
 }
 
+// ReadHeaderRange returns the rlp-encoded headers, starting at 'number', and going
+// backwards towards genesis. This method assumes that the caller already has
+// placed a cap on count, to prevent DoS issues.
+// Since this method operates in head-towards-genesis mode, it will return an empty
+// slice in case the head ('number') is missing. Hence, the caller must ensure that
+// the head ('number') argument is actually an existing header.
+//
+// N.B: Since the input is a number, as opposed to a hash, it's implicit that
+// this method only operates on canon headers.
+func ReadHeaderRange(db ethdb.Reader, number uint64, count uint64) []rlp.RawValue {
+	var rlpHeaders []rlp.RawValue
+	if count == 0 {
+		return rlpHeaders
+	}
+	i := number
+	if count-1 > number {
+		// It's ok to request block 0, 1 item
+		count = number + 1
+	}
+	limit, _ := db.Ancients()
+	// First read live blocks
+	if i >= limit {
+		// If we need to read live blocks, we need to figure out the hash first
+		hash := ReadCanonicalHash(db, number)
+		for ; i >= limit && count > 0; i-- {
+			if data, _ := db.Get(headerKey(i, hash)); len(data) > 0 {
+				rlpHeaders = append(rlpHeaders, data)
+				// Get the parent hash for next query
+				hash = types.HeaderParentHashFromRLP(data)
+			} else {
+				break // Maybe got moved to ancients
+			}
+			count--
+		}
+	}
+	if count == 0 {
+		return rlpHeaders
+	}
+	// read remaining from ancients
+	max := count * 700
+	data, err := db.AncientRange(freezerHeaderTable, i+1-count, count, max)
+	if err == nil && uint64(len(data)) == count {
+		// the data is on the order [h, h+1, .., n] -- reordering needed
+		for i := range data {
+			rlpHeaders = append(rlpHeaders, data[len(data)-1-i])
+		}
+	}
+	return rlpHeaders
+}
+
 // ReadHeaderRLP retrieves a block header in its raw RLP database encoding.
 func ReadHeaderRLP(db ethdb.Reader, hash common.Hash, number uint64) rlp.RawValue {
 	var data []byte

core/rawdb/accessors_chain_test.go

@@ -883,3 +883,67 @@ func BenchmarkDecodeRLPLogs(b *testing.B) {
 		}
 	})
 }
+
+func TestHeadersRLPStorage(t *testing.T) {
+	// Have N headers in the freezer
+	frdir, err := ioutil.TempDir("", "")
+	if err != nil {
+		t.Fatalf("failed to create temp freezer dir: %v", err)
+	}
+	defer os.Remove(frdir)
+
+	db, err := NewDatabaseWithFreezer(NewMemoryDatabase(), frdir, "", false)
+	if err != nil {
+		t.Fatalf("failed to create database with ancient backend")
+	}
+	defer db.Close()
+	// Create blocks
+	var chain []*types.Block
+	var pHash common.Hash
+	for i := 0; i < 100; i++ {
+		block := types.NewBlockWithHeader(&types.Header{
+			Number:      big.NewInt(int64(i)),
+			Extra:       []byte("test block"),
+			UncleHash:   types.EmptyUncleHash,
+			TxHash:      types.EmptyRootHash,
+			ReceiptHash: types.EmptyRootHash,
+			ParentHash:  pHash,
+		})
+		chain = append(chain, block)
+		pHash = block.Hash()
+	}
+	var receipts []types.Receipts = make([]types.Receipts, 100)
+	// Write first half to ancients
+	WriteAncientBlocks(db, chain[:50], receipts[:50], big.NewInt(100))
+	// Write second half to db
+	for i := 50; i < 100; i++ {
+		WriteCanonicalHash(db, chain[i].Hash(), chain[i].NumberU64())
+		WriteBlock(db, chain[i])
+	}
+	checkSequence := func(from, amount int) {
+		headersRlp := ReadHeaderRange(db, uint64(from), uint64(amount))
+		if have, want := len(headersRlp), amount; have != want {
+			t.Fatalf("have %d headers, want %d", have, want)
+		}
+		for i, headerRlp := range headersRlp {
+			var header types.Header
+			if err := rlp.DecodeBytes(headerRlp, &header); err != nil {
+				t.Fatal(err)
+			}
+			if have, want := header.Number.Uint64(), uint64(from-i); have != want {
+				t.Fatalf("wrong number, have %d want %d", have, want)
+			}
+		}
+	}
+	checkSequence(99, 20)  // Latest block and 19 parents
+	checkSequence(99, 50)  // Latest block -> all db blocks
+	checkSequence(99, 51)  // Latest block -> one from ancients
+	checkSequence(99, 52)  // Latest blocks -> two from ancients
+	checkSequence(50, 2)   // One from db, one from ancients
+	checkSequence(49, 1)   // One from ancients
+	checkSequence(49, 50)  // All ancient ones
+	checkSequence(99, 100) // All blocks
+	checkSequence(0, 1)    // Only genesis
+	checkSequence(1, 1)    // Only block 1
+	checkSequence(1, 2)    // Genesis + block 1
+}

core/types/block.go

@@ -389,3 +389,21 @@ func (b *Block) Hash() common.Hash {
 }
 
 type Blocks []*Block
+
+// HeaderParentHashFromRLP returns the parentHash of an RLP-encoded
+// header. If 'header' is invalid, the zero hash is returned.
+func HeaderParentHashFromRLP(header []byte) common.Hash {
+	// parentHash is the first list element.
+	listContent, _, err := rlp.SplitList(header)
+	if err != nil {
+		return common.Hash{}
+	}
+	parentHash, _, err := rlp.SplitString(listContent)
+	if err != nil {
+		return common.Hash{}
+	}
+	if len(parentHash) != 32 {
+		return common.Hash{}
+	}
+	return common.BytesToHash(parentHash)
+}

core/types/block_test.go

@@ -281,3 +281,64 @@ func makeBenchBlock() *Block {
 	}
 	return NewBlock(header, txs, uncles, receipts, newHasher())
 }
+
+func TestRlpDecodeParentHash(t *testing.T) {
+	// A minimum one
+	want := common.HexToHash("0x112233445566778899001122334455667788990011223344556677889900aabb")
+	if rlpData, err := rlp.EncodeToBytes(Header{ParentHash: want}); err != nil {
+		t.Fatal(err)
+	} else {
+		if have := HeaderParentHashFromRLP(rlpData); have != want {
+			t.Fatalf("have %x, want %x", have, want)
+		}
+	}
+	// And a maximum one
+	// | Difficulty  | dynamic| *big.Int       | 0x5ad3c2c71bbff854908 (current mainnet TD: 76 bits) |
+	// | Number      | dynamic| *big.Int       | 64 bits               |
+	// | Extra       | dynamic| []byte         | 65+32 byte (clique)   |
+	// | BaseFee     | dynamic| *big.Int       | 64 bits               |
+	mainnetTd := new(big.Int)
+	mainnetTd.SetString("5ad3c2c71bbff854908", 16)
+	if rlpData, err := rlp.EncodeToBytes(Header{
+		ParentHash: want,
+		Difficulty: mainnetTd,
+		Number:     new(big.Int).SetUint64(math.MaxUint64),
+		Extra:      make([]byte, 65+32),
+		BaseFee:    new(big.Int).SetUint64(math.MaxUint64),
+	}); err != nil {
+		t.Fatal(err)
+	} else {
+		if have := HeaderParentHashFromRLP(rlpData); have != want {
+			t.Fatalf("have %x, want %x", have, want)
+		}
+	}
+	// Also test a very very large header.
+	{
+		// The rlp-encoding of the heder belowCauses _total_ length of 65540,
+		// which is the first to blow the fast-path.
+		h := Header{
+			ParentHash: want,
+			Extra:      make([]byte, 65041),
+		}
+		if rlpData, err := rlp.EncodeToBytes(h); err != nil {
+			t.Fatal(err)
+		} else {
+			if have := HeaderParentHashFromRLP(rlpData); have != want {
+				t.Fatalf("have %x, want %x", have, want)
+			}
+		}
+	}
+	{
+		// Test some invalid erroneous stuff
+		for i, rlpData := range [][]byte{
+			nil,
+			common.FromHex("0x"),
+			common.FromHex("0x01"),
+			common.FromHex("0x3031323334"),
+		} {
+			if have, want := HeaderParentHashFromRLP(rlpData), (common.Hash{}); have != want {
+				t.Fatalf("invalid %d: have %x, want %x", i, have, want)
+			}
+		}
+	}
+}

eth/downloader/downloader_test.go

@@ -154,20 +154,32 @@ func (dlp *downloadTesterPeer) Head() (common.Hash, *big.Int) {
 	return head.Hash(), dlp.chain.GetTd(head.Hash(), head.NumberU64())
 }
 
+func unmarshalRlpHeaders(rlpdata []rlp.RawValue) []*types.Header {
+	var headers = make([]*types.Header, len(rlpdata))
+	for i, data := range rlpdata {
+		var h types.Header
+		if err := rlp.DecodeBytes(data, &h); err != nil {
+			panic(err)
+		}
+		headers[i] = &h
+	}
+	return headers
+}
+
 // RequestHeadersByHash constructs a GetBlockHeaders function based on a hashed
 // origin; associated with a particular peer in the download tester. The returned
 // function can be used to retrieve batches of headers from the particular peer.
 func (dlp *downloadTesterPeer) RequestHeadersByHash(origin common.Hash, amount int, skip int, reverse bool, sink chan *eth.Response) (*eth.Request, error) {
 	// Service the header query via the live handler code
-	headers := eth.ServiceGetBlockHeadersQuery(dlp.chain, &eth.GetBlockHeadersPacket{
+	rlpHeaders := eth.ServiceGetBlockHeadersQuery(dlp.chain, &eth.GetBlockHeadersPacket{
 		Origin: eth.HashOrNumber{
 			Hash: origin,
 		},
 		Amount:  uint64(amount),
 		Skip:    uint64(skip),
 		Reverse: reverse,
 	}, nil)
-
+	headers := unmarshalRlpHeaders(rlpHeaders)
 	// If a malicious peer is simulated withholding headers, delete them
 	for hash := range dlp.withholdHeaders {
 		for i, header := range headers {
@@ -203,15 +215,15 @@ func (dlp *downloadTesterPeer) RequestHeadersByHash(origin common.Hash, amount i
 // function can be used to retrieve batches of headers from the particular peer.
 func (dlp *downloadTesterPeer) RequestHeadersByNumber(origin uint64, amount int, skip int, reverse bool, sink chan *eth.Response) (*eth.Request, error) {
 	// Service the header query via the live handler code
-	headers := eth.ServiceGetBlockHeadersQuery(dlp.chain, &eth.GetBlockHeadersPacket{
+	rlpHeaders := eth.ServiceGetBlockHeadersQuery(dlp.chain, &eth.GetBlockHeadersPacket{
 		Origin: eth.HashOrNumber{
 			Number: origin,
 		},
 		Amount:  uint64(amount),
 		Skip:    uint64(skip),
 		Reverse: reverse,
 	}, nil)
-
+	headers := unmarshalRlpHeaders(rlpHeaders)
 	// If a malicious peer is simulated withholding headers, delete them
 	for hash := range dlp.withholdHeaders {
 		for i, header := range headers {

eth/handler_eth_test.go

@@ -38,6 +38,7 @@ import (
 	"github.com/ethereum/go-ethereum/p2p"
 	"github.com/ethereum/go-ethereum/p2p/enode"
 	"github.com/ethereum/go-ethereum/params"
+	"github.com/ethereum/go-ethereum/rlp"
 )
 
 // testEthHandler is a mock event handler to listen for inbound network requests
@@ -560,15 +561,17 @@ func testCheckpointChallenge(t *testing.T, syncmode downloader.SyncMode, checkpo
 		// Create a block to reply to the challenge if no timeout is simulated.
 		if !timeout {
 			if empty {
-				if err := remote.ReplyBlockHeaders(request.RequestId, []*types.Header{}); err != nil {
+				if err := remote.ReplyBlockHeadersRLP(request.RequestId, []rlp.RawValue{}); err != nil {
 					t.Fatalf("failed to answer challenge: %v", err)
 				}
 			} else if match {
-				if err := remote.ReplyBlockHeaders(request.RequestId, []*types.Header{response}); err != nil {
+				responseRlp, _ := rlp.EncodeToBytes(response)
+				if err := remote.ReplyBlockHeadersRLP(request.RequestId, []rlp.RawValue{responseRlp}); err != nil {
 					t.Fatalf("failed to answer challenge: %v", err)
 				}
 			} else {
-				if err := remote.ReplyBlockHeaders(request.RequestId, []*types.Header{{Number: response.Number}}); err != nil {
+				responseRlp, _ := rlp.EncodeToBytes(types.Header{Number: response.Number})
+				if err := remote.ReplyBlockHeadersRLP(request.RequestId, []rlp.RawValue{responseRlp}); err != nil {
 					t.Fatalf("failed to answer challenge: %v", err)
 				}
 			}

eth/protocols/eth/handler.go

@@ -35,9 +35,6 @@ const (
 	// softResponseLimit is the target maximum size of replies to data retrievals.
 	softResponseLimit = 2 * 1024 * 1024
 
-	// estHeaderSize is the approximate size of an RLP encoded block header.
-	estHeaderSize = 500
-
 	// maxHeadersServe is the maximum number of block headers to serve. This number
 	// is there to limit the number of disk lookups.
 	maxHeadersServe = 1024