PARTIAL sync: coreth PR #963,981,1009: nodeType and related items

plugin/evm/extension/config.go

@@ -10,6 +10,7 @@ import (
 
 	"github.com/ava-labs/subnet-evm/plugin/evm/message"
 	"github.com/ava-labs/subnet-evm/plugin/evm/sync"
+	"github.com/ava-labs/subnet-evm/sync/handlers"
 )
 
 var (
@@ -19,6 +20,17 @@ var (
 	errNilClock               = errors.New("nil clock")
 )
 
+// LeafRequestConfig is the configuration to handle leaf requests
+// in the network and syncer
+type LeafRequestConfig struct {
+	// LeafType is the type of the leaf node
+	LeafType message.NodeType
+	// MetricName is the name of the metric to use for the leaf request
+	MetricName string
+	// Handler is the handler to use for the leaf request
+	Handler handlers.LeafRequestHandler
+}
+
 // Config is the configuration for the VM extension
 type Config struct {
 	// SyncSummaryProvider is the sync summary provider to use

plugin/evm/message/handler.go

@@ -15,9 +15,8 @@ var _ RequestHandler = NoopRequestHandler{}
 // Must have methods in format of handleType(context.Context, ids.NodeID, uint32, request Type) error
 // so that the Request object of relevant Type can invoke its respective handle method
 // on this struct.
-// Also see GossipHandler for implementation style.
 type RequestHandler interface {
-	HandleStateTrieLeafsRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, leafsRequest LeafsRequest) ([]byte, error)
+	HandleLeafsRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, leafsRequest LeafsRequest) ([]byte, error)
 	HandleBlockRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, request BlockRequest) ([]byte, error)
 	HandleCodeRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, codeRequest CodeRequest) ([]byte, error)
 }
@@ -33,7 +32,7 @@ type ResponseHandler interface {
 
 type NoopRequestHandler struct{}
 
-func (NoopRequestHandler) HandleStateTrieLeafsRequest(context.Context, ids.NodeID, uint32, LeafsRequest) ([]byte, error) {
+func (NoopRequestHandler) HandleLeafsRequest(context.Context, ids.NodeID, uint32, LeafsRequest) ([]byte, error) {
 	return nil, nil
 }
 

plugin/evm/message/leafs_request.go

@@ -13,27 +13,37 @@ import (
 
 const MaxCodeHashesPerRequest = 5
 
-var _ Request = LeafsRequest{}
+// NodeType outlines the trie that a leaf node belongs to
+// handlers.LeafsRequestHandler uses this information to determine
+// which trie type to fetch the information from
+type NodeType uint8
+
+const (
+	StateTrieNode      = NodeType(1)
+	StateTrieKeyLength = common.HashLength
+)
 
 // LeafsRequest is a request to receive trie leaves at specified Root within Start and End byte range
 // Limit outlines maximum number of leaves to returns starting at Start
+// NodeType outlines which trie to read from state/atomic.
 type LeafsRequest struct {
-	Root    common.Hash `serialize:"true"`
-	Account common.Hash `serialize:"true"`
-	Start   []byte      `serialize:"true"`
-	End     []byte      `serialize:"true"`
-	Limit   uint16      `serialize:"true"`
+	Root     common.Hash `serialize:"true"`
+	Account  common.Hash `serialize:"true"`
+	Start    []byte      `serialize:"true"`
+	End      []byte      `serialize:"true"`
+	Limit    uint16      `serialize:"true"`
+	NodeType NodeType    `serialize:"true"`
 }
 
 func (l LeafsRequest) String() string {
 	return fmt.Sprintf(
-		"LeafsRequest(Root=%s, Account=%s, Start=%s, End %s, Limit=%d)",
-		l.Root, l.Account, common.Bytes2Hex(l.Start), common.Bytes2Hex(l.End), l.Limit,
+		"LeafsRequest(Root=%s, Account=%s, Start=%s, End=%s, Limit=%d, NodeType=%d)",
+		l.Root, l.Account, common.Bytes2Hex(l.Start), common.Bytes2Hex(l.End), l.Limit, l.NodeType,
 	)
 }
 
 func (l LeafsRequest) Handle(ctx context.Context, nodeID ids.NodeID, requestID uint32, handler RequestHandler) ([]byte, error) {
-	return handler.HandleStateTrieLeafsRequest(ctx, nodeID, requestID, l)
+	return handler.HandleLeafsRequest(ctx, nodeID, requestID, l)
 }
 
 // LeafsResponse is a response to a LeafsRequest

plugin/evm/message/leafs_request_test.go

@@ -4,13 +4,10 @@
 package message
 
 import (
-	"bytes"
-	"context"
 	"encoding/base64"
 	"math/rand"
 	"testing"
 
-	"github.com/ava-labs/avalanchego/ids"
 	"github.com/ava-labs/libevm/common"
 	"github.com/stretchr/testify/assert"
 )
@@ -38,7 +35,7 @@ func TestMarshalLeafsRequest(t *testing.T) {
 		Limit: 1024,
 	}
 
-	base64LeafsRequest := "AAAAAAAAAAAAAAAAAAAAAABpbSBST09UaW5nIGZvciB5YQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAIFL9/AchgmVPFj9fD5piHXKVZsdNEAN8TXu7BAfR4sZJAAAAIIGFWthoHQ2G0ekeABZ5OctmlNLEIqzSCKAHKTlIf2mZBAA="
+	base64LeafsRequest := "AAAAAAAAAAAAAAAAAAAAAABpbSBST09UaW5nIGZvciB5YQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAIFL9/AchgmVPFj9fD5piHXKVZsdNEAN8TXu7BAfR4sZJAAAAIIGFWthoHQ2G0ekeABZ5OctmlNLEIqzSCKAHKTlIf2mZBAAA"
 
 	leafsRequestBytes, err := Codec.Marshal(Version, leafsRequest)
 	assert.NoError(t, err)
@@ -105,62 +102,3 @@ func TestMarshalLeafsResponse(t *testing.T) {
 	assert.False(t, l.More) // make sure it is not serialized
 	assert.Equal(t, leafsResponse.ProofVals, l.ProofVals)
 }
-
-func TestLeafsRequestValidation(t *testing.T) {
-	mockRequestHandler := &mockHandler{}
-
-	tests := map[string]struct {
-		request        LeafsRequest
-		assertResponse func(t *testing.T)
-	}{
-		"node type StateTrieNode": {
-			request: LeafsRequest{
-				Root:  common.BytesToHash([]byte("some hash goes here")),
-				Start: bytes.Repeat([]byte{0x00}, common.HashLength),
-				End:   bytes.Repeat([]byte{0xff}, common.HashLength),
-				Limit: 10,
-			},
-			assertResponse: func(t *testing.T) {
-				assert.True(t, mockRequestHandler.handleStateTrieCalled)
-				assert.False(t, mockRequestHandler.handleBlockRequestCalled)
-				assert.False(t, mockRequestHandler.handleCodeRequestCalled)
-			},
-		},
-	}
-	for name, test := range tests {
-		t.Run(name, func(t *testing.T) {
-			_, _ = test.request.Handle(context.Background(), ids.GenerateTestNodeID(), 1, mockRequestHandler)
-			test.assertResponse(t)
-			mockRequestHandler.reset()
-		})
-	}
-}
-
-var _ RequestHandler = (*mockHandler)(nil)
-
-type mockHandler struct {
-	handleStateTrieCalled,
-	handleBlockRequestCalled,
-	handleCodeRequestCalled bool
-}
-
-func (m *mockHandler) HandleStateTrieLeafsRequest(context.Context, ids.NodeID, uint32, LeafsRequest) ([]byte, error) {
-	m.handleStateTrieCalled = true
-	return nil, nil
-}
-
-func (m *mockHandler) HandleBlockRequest(context.Context, ids.NodeID, uint32, BlockRequest) ([]byte, error) {
-	m.handleBlockRequestCalled = true
-	return nil, nil
-}
-
-func (m *mockHandler) HandleCodeRequest(context.Context, ids.NodeID, uint32, CodeRequest) ([]byte, error) {
-	m.handleCodeRequestCalled = true
-	return nil, nil
-}
-
-func (m *mockHandler) reset() {
-	m.handleStateTrieCalled = false
-	m.handleBlockRequestCalled = false
-	m.handleCodeRequestCalled = false
-}

plugin/evm/network_handler.go

@@ -9,7 +9,7 @@ import (
 	"github.com/ava-labs/avalanchego/codec"
 	"github.com/ava-labs/avalanchego/ids"
 	"github.com/ava-labs/libevm/ethdb"
-	"github.com/ava-labs/libevm/metrics"
+	"github.com/ava-labs/libevm/log"
 	"github.com/ava-labs/libevm/triedb"
 
 	"github.com/ava-labs/subnet-evm/plugin/evm/message"
@@ -20,29 +20,44 @@ import (
 
 var _ message.RequestHandler = (*networkHandler)(nil)
 
+type LeafHandlers map[message.NodeType]syncHandlers.LeafRequestHandler
+
 type networkHandler struct {
-	leafRequestHandler  *syncHandlers.LeafsRequestHandler
+	leafRequestHandlers LeafHandlers
 	blockRequestHandler *syncHandlers.BlockRequestHandler
 	codeRequestHandler  *syncHandlers.CodeRequestHandler
 }
 
+type LeafRequestTypeConfig struct {
+	NodeType     message.NodeType
+	NodeKeyLen   int
+	TrieDB       *triedb.Database
+	UseSnapshots bool
+	MetricName   string
+}
+
 // newNetworkHandler constructs the handler for serving network requests.
 func newNetworkHandler(
 	provider syncHandlers.SyncDataProvider,
 	diskDB ethdb.KeyValueReader,
-	evmTrieDB *triedb.Database,
 	networkCodec codec.Manager,
-) message.RequestHandler {
-	syncStats := syncStats.NewHandlerStats(metrics.Enabled)
+	leafRequestHandlers LeafHandlers,
+	syncStats syncStats.HandlerStats,
+) *networkHandler {
 	return &networkHandler{
-		leafRequestHandler:  syncHandlers.NewLeafsRequestHandler(evmTrieDB, nil, networkCodec, syncStats),
+		leafRequestHandlers: leafRequestHandlers,
 		blockRequestHandler: syncHandlers.NewBlockRequestHandler(provider, networkCodec, syncStats),
 		codeRequestHandler:  syncHandlers.NewCodeRequestHandler(diskDB, networkCodec, syncStats),
 	}
 }
 
-func (n networkHandler) HandleStateTrieLeafsRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, leafsRequest message.LeafsRequest) ([]byte, error) {
-	return n.leafRequestHandler.OnLeafsRequest(ctx, nodeID, requestID, leafsRequest)
+func (n networkHandler) HandleLeafsRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, leafsRequest message.LeafsRequest) ([]byte, error) {
+	handler, ok := n.leafRequestHandlers[leafsRequest.NodeType]
+	if !ok {
+		log.Debug("node type is not recognised, dropping request", "nodeID", nodeID, "requestID", requestID, "nodeType", leafsRequest.NodeType)
+		return nil, nil
+	}
+	return handler.OnLeafsRequest(ctx, nodeID, requestID, leafsRequest)
 }
 
 func (n networkHandler) HandleBlockRequest(ctx context.Context, nodeID ids.NodeID, requestID uint32, blockRequest message.BlockRequest) ([]byte, error) {

plugin/evm/vm.go

@@ -76,6 +76,7 @@ import (
 	"github.com/ava-labs/subnet-evm/precompile/precompileconfig"
 	"github.com/ava-labs/subnet-evm/rpc"
 	"github.com/ava-labs/subnet-evm/sync/client/stats"
+	"github.com/ava-labs/subnet-evm/sync/handlers"
 	"github.com/ava-labs/subnet-evm/triedb/hashdb"
 	"github.com/ava-labs/subnet-evm/warp"
 
@@ -88,6 +89,7 @@ import (
 	subnetevmlog "github.com/ava-labs/subnet-evm/plugin/evm/log"
 	vmsync "github.com/ava-labs/subnet-evm/plugin/evm/sync"
 	statesyncclient "github.com/ava-labs/subnet-evm/sync/client"
+	handlerstats "github.com/ava-labs/subnet-evm/sync/handlers/stats"
 	avalancheRPC "github.com/gorilla/rpc/v2"
 )
 
@@ -503,11 +505,9 @@ func (vm *VM) Initialize(
 	warpHandler := acp118.NewCachedHandler(meteredCache, vm.warpBackend, vm.ctx.WarpSigner)
 	vm.Network.AddHandler(p2p.SignatureRequestHandlerID, warpHandler)
 
-	vm.setAppRequestHandlers()
-
 	vm.stateSyncDone = make(chan struct{})
 
-	return vm.initializeStateSyncClient(lastAcceptedHeight)
+	return vm.initializeStateSync(lastAcceptedHeight)
 }
 
 func parseGenesis(ctx *snow.Context, genesisBytes []byte, upgradeBytes []byte, airdropFile string) (*core.Genesis, error) {
@@ -651,7 +651,44 @@ func (vm *VM) initializeChain(lastAcceptedHash common.Hash, ethConfig ethconfig.
 // initializeStateSyncClient initializes the client for performing state sync.
 // If state sync is disabled, this function will wipe any ongoing summary from
 // disk to ensure that we do not continue syncing from an invalid snapshot.
-func (vm *VM) initializeStateSyncClient(lastAcceptedHeight uint64) error {
+func (vm *VM) initializeStateSync(lastAcceptedHeight uint64) error {
+	// Create standalone EVM TrieDB (read only) for serving leafs requests.
+	// We create a standalone TrieDB here, so that it has a standalone cache from the one
+	// used by the node when processing blocks.
+	evmTrieDB := triedb.NewDatabase(
+		vm.chaindb,
+		&triedb.Config{
+			DBOverride: hashdb.Config{
+				CleanCacheSize: vm.config.StateSyncServerTrieCache * units.MiB,
+			}.BackendConstructor,
+		},
+	)
+
+	// register default leaf request handler for state trie
+	syncStats := handlerstats.GetOrRegisterHandlerStats(metrics.Enabled)
+	stateLeafRequestConfig := &extension.LeafRequestConfig{
+		LeafType:   message.StateTrieNode,
+		MetricName: "sync_state_trie_leaves",
+		Handler: handlers.NewLeafsRequestHandler(evmTrieDB,
+			message.StateTrieKeyLength,
+			vm.blockChain, vm.networkCodec,
+			syncStats,
+		),
+	}
+
+	leafHandlers := make(LeafHandlers)
+	leafHandlers[stateLeafRequestConfig.LeafType] = stateLeafRequestConfig.Handler
+
+	networkHandler := newNetworkHandler(
+		vm.blockChain,
+		vm.chaindb,
+		vm.networkCodec,
+		leafHandlers,
+		syncStats,
+	)
+	vm.Network.SetRequestHandler(networkHandler)
+
+	vm.Server = vmsync.NewServer(vm.blockChain, vm.extensionConfig.SyncSummaryProvider, vm.config.StateSyncCommitInterval) // parse nodeIDs from state sync IDs in vm config
 	// parse nodeIDs from state sync IDs in vm config
 	var stateSyncIDs []ids.NodeID
 	if vm.config.StateSyncEnabled && len(vm.config.StateSyncIDs) > 0 {
@@ -666,15 +703,19 @@ func (vm *VM) initializeStateSyncClient(lastAcceptedHeight uint64) error {
 		}
 	}
 
+	// Initialize the state sync client
+	leafMetricsNames := make(map[message.NodeType]string)
+	leafMetricsNames[stateLeafRequestConfig.LeafType] = stateLeafRequestConfig.MetricName
+
 	vm.Client = vmsync.NewClient(&vmsync.ClientConfig{
+		StateSyncDone: vm.stateSyncDone,
 		Chain:         vm.eth,
 		State:         vm.State,
-		StateSyncDone: vm.stateSyncDone,
 		Client: statesyncclient.NewClient(
 			&statesyncclient.ClientConfig{
 				NetworkClient:    vm.Network,
 				Codec:            vm.networkCodec,
-				Stats:            stats.NewClientSyncerStats(),
+				Stats:            stats.NewClientSyncerStats(leafMetricsNames),
 				StateSyncNodeIDs: stateSyncIDs,
 				BlockParser:      vm,
 			},
@@ -895,27 +936,6 @@ func (vm *VM) onNormalOperationsStarted() error {
 	return nil
 }
 
-// setAppRequestHandlers sets the request handlers for the VM to serve state sync
-// requests.
-func (vm *VM) setAppRequestHandlers() {
-	// Create standalone EVM TrieDB (read only) for serving leafs requests.
-	// We create a standalone TrieDB here, so that it has a standalone cache from the one
-	// used by the node when processing blocks.
-	evmTrieDB := triedb.NewDatabase(
-		vm.chaindb,
-		&triedb.Config{
-			DBOverride: hashdb.Config{
-				CleanCacheSize: vm.config.StateSyncServerTrieCache * units.MiB,
-			}.BackendConstructor,
-		},
-	)
-
-	networkHandler := newNetworkHandler(vm.blockChain, vm.chaindb, evmTrieDB, vm.networkCodec)
-	vm.Network.SetRequestHandler(networkHandler)
-
-	vm.Server = vmsync.NewServer(vm.blockChain, vm.extensionConfig.SyncSummaryProvider, vm.config.StateSyncCommitInterval)
-}
-
 func (vm *VM) WaitForEvent(ctx context.Context) (commonEng.Message, error) {
 	vm.builderLock.Lock()
 	builder := vm.builder

sync/README.md

@@ -42,7 +42,7 @@ When a new node wants to join the network via state sync, it will need a few pie
 - Number (height) and hash of the latest available syncable block,
 - Root of the account trie,
 
-The above information is called a _state summary_, and each syncable block corresponds to one such summary (see `message.SyncSummary`). The engine and VM interact as follows to find a syncable state summary:
+The above information is called a _state summary_, and each syncable block corresponds to one such summary (see `message.Summary`). The engine and VM interact as follows to find a syncable state summary:
 
 
 1. The engine calls `StateSyncEnabled`. The VM returns `true` to initiate state sync, or `false` to start  bootstrapping. In `subnet-evm`, this is controlled by the `state-sync-enabled` flag.
@@ -60,6 +60,8 @@ The following steps are executed by the VM to sync its state from peers (see `st
 1. Update in-memory and on-disk pointers.
 
 Steps 3 and 4 involve syncing tries. To sync trie data, the VM will send a series of `LeafRequests` to its peers. Each request specifies:
+- Type of trie (`NodeType`):
+  - `statesync.StateTrieNode` (account trie and storage tries share the same database)
 - `Root` of the trie to sync,
 - `Start` and `End` specify a range of keys.
 
@@ -112,4 +114,4 @@ While state sync is faster than normal bootstrapping, the process may take sever
 | `state-sync-skip-resume` | `bool` | set to true to avoid resuming an ongoing sync | `false` |
 | `state-sync-min-blocks` | `uint64` | Minimum number of blocks the chain must be ahead of local state to prefer state sync over bootstrapping | `300,000` |
 | `state-sync-server-trie-cache` | `int` | Size of trie cache to serve state sync data in MB. Should be set to multiples of `64`. | `64` |
-| `state-sync-ids` | `string` | a comma separated list of `NodeID-` prefixed node IDs to sync data from. If not provided, peers are randomly selected. | |
+| `state-sync-ids` | `string` | a comma separated list of `NodeID-` prefixed node IDs to sync data from. If not provided, peers are randomly selected. | |