etcdserver: add --max-request-bytes flag

embed/config.go

@@ -36,14 +36,13 @@ import (
 const (
 	ClusterStateFlagNew      = "new"
 	ClusterStateFlagExisting = "existing"
-
-	DefaultName         = "default"
-	DefaultMaxSnapshots = 5
-	DefaultMaxWALs      = 5
-	DefaultMaxTxnOps    = uint(128)
-
-	DefaultListenPeerURLs   = "http://localhost:2380"
-	DefaultListenClientURLs = "http://localhost:2379"
+	DefaultName              = "default"
+	DefaultMaxSnapshots      = 5
+	DefaultMaxWALs           = 5
+	DefaultMaxTxnOps         = uint(128)
+	DefaultMaxRequestBytes   = 1.5 * 1024 * 1024
+	DefaultListenPeerURLs    = "http://localhost:2380"
+	DefaultListenClientURLs  = "http://localhost:2379"
 
 	// maxElectionMs specifies the maximum value of election timeout.
 	// More details are listed in ../Documentation/tuning.md#time-parameters.
@@ -87,7 +86,8 @@ type Config struct {
 	ElectionMs        uint  `json:"election-timeout"`
 	QuotaBackendBytes int64 `json:"quota-backend-bytes"`
 	MaxTxnOps         uint  `json:"max-txn-ops"`
-
+	MaxRequestBytes   uint  `json:"max-request-bytes"`
+
 	// clustering
 
 	APUrls, ACUrls      []url.URL
@@ -175,6 +175,7 @@ func NewConfig() *Config {
 		Name:                DefaultName,
 		SnapCount:           etcdserver.DefaultSnapCount,
 		MaxTxnOps:           DefaultMaxTxnOps,
+		MaxRequestBytes:     DefaultMaxRequestBytes,
 		TickMs:              100,
 		ElectionMs:          1000,
 		LPUrls:              []url.URL{*lpurl},

embed/etcd.go

@@ -140,6 +140,7 @@ func StartEtcd(inCfg *Config) (e *Etcd, err error) {
 		AutoCompactionRetention: cfg.AutoCompactionRetention,
 		QuotaBackendBytes:       cfg.QuotaBackendBytes,
 		MaxTxnOps:               cfg.MaxTxnOps,
+		MaxRequestBytes:         cfg.MaxRequestBytes,
 		StrictReconfigCheck:     cfg.StrictReconfigCheck,
 		ClientCertAuthEnabled:   cfg.ClientTLSInfo.ClientCertAuth,
 		AuthToken:               cfg.AuthToken,

etcdmain/config.go

@@ -139,6 +139,7 @@ func newConfig() *config {
 	fs.UintVar(&cfg.ElectionMs, "election-timeout", cfg.ElectionMs, "Time (in milliseconds) for an election to timeout.")
 	fs.Int64Var(&cfg.QuotaBackendBytes, "quota-backend-bytes", cfg.QuotaBackendBytes, "Raise alarms when backend size exceeds the given quota. 0 means use the default quota.")
 	fs.UintVar(&cfg.MaxTxnOps, "max-txn-ops", cfg.MaxTxnOps, "Maximum number of operations permitted in a transaction.")
+	fs.UintVar(&cfg.MaxRequestBytes, "max-request-bytes", cfg.MaxRequestBytes, "Maximum client request size in bytes the server will accept.")
 
 	// clustering
 	fs.Var(flags.NewURLsValue(embed.DefaultInitialAdvertisePeerURLs), "initial-advertise-peer-urls", "List of this member's peer URLs to advertise to the rest of the cluster.")

etcdmain/help.go

@@ -68,6 +68,8 @@ member flags:
 		raise alarms when backend size exceeds the given quota (0 defaults to low space quota).
 	--max-txn-ops '128' 
 		maximum number of operations permitted in a transaction.
+	--max-request-bytes '1572864'
+		maximum client request size in bytes the server will accept.
 
 clustering flags:
 

etcdserver/config.go

@@ -56,6 +56,9 @@ type ServerConfig struct {
 	QuotaBackendBytes       int64
 	MaxTxnOps               uint
 
+	// MaxRequestBytes is the maximum request size to send over raft.
+	MaxRequestBytes uint
+
 	StrictReconfigCheck bool
 
 	// ClientCertAuthEnabled is true when cert has been signed by the client CA.

etcdserver/server.go

@@ -83,6 +83,8 @@ const (
 
 	// maxPendingRevokes is the maximum number of outstanding expired lease revocations.
 	maxPendingRevokes = 16
+
+	recommendedMaxRequestBytes = 10 * 1024 * 1024
 )
 
 var (
@@ -259,6 +261,10 @@ func NewServer(cfg *ServerConfig) (srv *EtcdServer, err error) {
 		cl *membership.RaftCluster
 	)
 
+	if cfg.MaxRequestBytes > recommendedMaxRequestBytes {
+		plog.Warningf("MaxRequestBytes %v exceeds maximum recommended size %v", cfg.MaxRequestBytes, recommendedMaxRequestBytes)
+	}
+
 	if terr := fileutil.TouchDirAll(cfg.DataDir); terr != nil {
 		return nil, fmt.Errorf("cannot access data directory: %v", terr)
 	}

etcdserver/v3_server.go

@@ -31,12 +31,6 @@ import (
 )
 
 const (
-	// the max request size that raft accepts.
-	// TODO: make this a flag? But we probably do not want to
-	// accept large request which might block raft stream. User
-	// specify a large value might end up with shooting in the foot.
-	maxRequestBytes = 1.5 * 1024 * 1024
-
 	// In the health case, there might be a small gap (10s of entries) between
 	// the applied index and committed index.
 	// However, if the committed entries are very heavy to apply, the gap might grow.
@@ -605,7 +599,7 @@ func (s *EtcdServer) processInternalRaftRequestOnce(ctx context.Context, r pb.In
 		return nil, err
 	}
 
-	if len(data) > maxRequestBytes {
+	if len(data) > int(s.Cfg.MaxRequestBytes) {
 		return nil, ErrRequestTooLarge
 	}
 

integration/cluster.go

@@ -61,6 +61,8 @@ const (
 	basePort     = 21000
 	UrlScheme    = "unix"
 	UrlSchemeTLS = "unixs"
+
+	defaultMaxRequestSize
 )
 
 var (
@@ -95,6 +97,7 @@ type ClusterConfig struct {
 	UseGRPC           bool
 	QuotaBackendBytes int64
 	MaxTxnOps         uint
+	MaxRequestBytes   uint
 }
 
 type cluster struct {
@@ -227,6 +230,7 @@ func (c *cluster) mustNewMember(t *testing.T) *member {
 			clientTLS:         c.cfg.ClientTLS,
 			quotaBackendBytes: c.cfg.QuotaBackendBytes,
 			maxTxnOps:         c.cfg.MaxTxnOps,
+			maxRequestBytes:   c.cfg.MaxRequestBytes,
 		})
 	m.DiscoveryURL = c.cfg.DiscoveryURL
 	if c.cfg.UseGRPC {
@@ -494,6 +498,7 @@ type memberConfig struct {
 	clientTLS         *transport.TLSInfo
 	quotaBackendBytes int64
 	maxTxnOps         uint
+	maxRequestBytes   uint
 }
 
 // mustNewMember return an inited member with the given name. If peerTLS is
@@ -545,6 +550,10 @@ func mustNewMember(t *testing.T, mcfg memberConfig) *member {
 	if m.MaxTxnOps == 0 {
 		m.MaxTxnOps = embed.DefaultMaxTxnOps
 	}
+	m.MaxRequestBytes = mcfg.maxRequestBytes
+	if m.MaxRequestBytes == 0 {
+		m.MaxRequestBytes = embed.DefaultMaxRequestBytes
+	}
 	m.AuthToken = "simple" // for the purpose of integration testing, simple token is enough
 	return m
 }

integration/v3_grpc_test.go

@@ -1625,6 +1625,35 @@ func TestGRPCStreamRequireLeader(t *testing.T) {
 	}
 }
 
+// TestV3PutLargeRequests ensures that configurable MaxRequestBytes works as intended.
+func TestV3PutLargeRequests(t *testing.T) {
+	defer testutil.AfterTest(t)
+	tests := []struct {
+		key             string
+		maxRequestBytes uint
+		valueSize       int
+		expectError     error
+	}{
+		// don't set to 0. use 0 as the default.
+		{"foo", 1, 1024, rpctypes.ErrGRPCRequestTooLarge},
+		{"foo", 10 * 1024 * 1024, 9 * 1024 * 1024, nil},
+		{"foo", 10 * 1024 * 1024, 10 * 1024 * 1024, rpctypes.ErrGRPCRequestTooLarge},
+		{"foo", 10 * 1024 * 1024, 10*1024*1024 + 5, rpctypes.ErrGRPCRequestTooLarge},
+	}
+	for i, test := range tests {
+		clus := NewClusterV3(t, &ClusterConfig{Size: 1, MaxRequestBytes: test.maxRequestBytes})
+		kvcli := toGRPC(clus.Client(0)).KV
+		reqput := &pb.PutRequest{Key: []byte(test.key), Value: make([]byte, test.valueSize)}
+		_, err := kvcli.Put(context.TODO(), reqput)
+
+		if !eqErrGRPC(err, test.expectError) {
+			t.Errorf("#%d: expected error %v, got %v", i, test.expectError, err)
+		}
+
+		clus.Terminate(t)
+	}
+}
+
 func eqErrGRPC(err1 error, err2 error) bool {
 	return !(err1 == nil && err2 != nil) || err1.Error() == err2.Error()
 }