fix: panic on cluster redirection to a node with stale role

Signed-off-by: Rueian <[email protected]>
redis · Dec 12, 2024 · bd347fd · bd347fd
1 parent 723761f
commit bd347fd
Show file tree

Hide file tree

Showing 2 changed files with 18 additions and 100 deletions.
diff --git a/cluster.go b/cluster.go
@@ -37,9 +37,9 @@ type clusterClient struct {
 
 // NOTE: connrole and conn must be initialized at the same time
 type connrole struct {
-	conn    conn
-	replica bool
-	hidden  bool
+	conn   conn
+	hidden bool
+	//replica bool <- this field is removed because a server may have mixed roles at the same time in the future. https://github.com/valkey-io/valkey/issues/1372
 }
 
 func newClusterClient(opt *ClientOption, connFn connFn, retryer retryHandler) (*clusterClient, error) {
@@ -192,12 +192,14 @@ func (c *clusterClient) _refresh() (err error) {
 	groups := result.parse(c.opt.TLSConfig != nil)
 	conns := make(map[string]connrole, len(groups))
 	for master, g := range groups {
-		conns[master] = connrole{conn: c.connFn(master, c.opt), replica: false}
-		for _, addr := range g.nodes[1:] {
-			if c.rOpt != nil {
-				conns[addr] = connrole{conn: c.connFn(addr, c.rOpt), replica: true}
-			} else {
-				conns[addr] = connrole{conn: c.connFn(addr, c.opt), replica: true}
+		conns[master] = connrole{conn: c.connFn(master, c.opt)}
+		if c.rOpt != nil {
+			for _, addr := range g.nodes[1:] {
+				conns[addr] = connrole{conn: c.connFn(addr, c.rOpt)}
+			}
+		} else {
+			for _, addr := range g.nodes[1:] {
+				conns[addr] = connrole{conn: c.connFn(addr, c.opt)}
 			}
 		}
 	}
@@ -215,13 +217,9 @@ func (c *clusterClient) _refresh() (err error) {
 
 	c.mu.RLock()
 	for addr, cc := range c.conns {
-		fresh, ok := conns[addr]
-		if ok && (cc.replica == fresh.replica || c.rOpt == nil) {
-			conns[addr] = connrole{
-				conn:    cc.conn,
-				replica: fresh.replica,
-				hidden:  fresh.hidden,
-			}
+		if fresh, ok := conns[addr]; ok {
+			fresh.conn = cc.conn
+			conns[addr] = fresh
 		} else {
 			removes = append(removes, cc.conn)
 		}
@@ -397,9 +395,6 @@ func (c *clusterClient) _pick(slot uint16, toReplica bool) (p conn) {
 	c.mu.RLock()
 	if slot == cmds.InitSlot {
 		for _, cc := range c.conns {
-			if cc.replica {
-				continue
-			}
 			p = cc.conn
 			break
 		}
@@ -434,7 +429,7 @@ func (c *clusterClient) redirectOrNew(addr string, prev conn, slot uint16, mode
 	c.mu.Lock()
 	if cc = c.conns[addr]; cc.conn == nil {
 		p := c.connFn(addr, c.opt)
-		cc = connrole{conn: p, replica: false}
+		cc = connrole{conn: p}
 		c.conns[addr] = cc
 		if mode == RedirectMove {
 			c.pslots[slot] = p
@@ -448,14 +443,10 @@ func (c *clusterClient) redirectOrNew(addr string, prev conn, slot uint16, mode
 			prev.Close()
 		}(prev)
 		p := c.connFn(addr, c.opt)
-		cc = connrole{conn: p, replica: cc.replica}
+		cc = connrole{conn: p}
 		c.conns[addr] = cc
-		if mode == RedirectMove {
-			if cc.replica {
-				c.rslots[slot] = p
-			} else {
-				c.pslots[slot] = p
-			}
+		if mode == RedirectMove { // MOVED should always point to the primary.
+			c.pslots[slot] = p
 		}
 	}
 	c.mu.Unlock()

diff --git a/cluster_test.go b/cluster_test.go
@@ -33,23 +33,6 @@ var slotsResp = newResult(RedisMessage{typ: '*', values: []RedisMessage{
 	}},
 }}, nil)
 
-var slotsRespWithChangedRole = newResult(RedisMessage{typ: '*', values: []RedisMessage{
-	{typ: '*', values: []RedisMessage{
-		{typ: ':', integer: 0},
-		{typ: ':', integer: 16383},
-		{typ: '*', values: []RedisMessage{ // master
-			{typ: '+', string: "127.0.1.1"},
-			{typ: ':', integer: 1},
-			{typ: '+', string: ""},
-		}},
-		{typ: '*', values: []RedisMessage{ // replica
-			{typ: '+', string: "127.0.0.1"},
-			{typ: ':', integer: 0},
-			{typ: '+', string: ""},
-		}},
-	}},
-}}, nil)
-
 var slotsMultiResp = newResult(RedisMessage{typ: '*', values: []RedisMessage{
 	{typ: '*', values: []RedisMessage{
 		{typ: ':', integer: 0},
@@ -1937,13 +1920,6 @@ func TestClusterClient_SendToOnlyPrimaryNodes(t *testing.T) {
 		t.Fatalf("unexpected err %v", err)
 	}
 
-	t.Run("Do with no slot", func(t *testing.T) {
-		c := client.B().Info().Build()
-		if v, err := client.Do(context.Background(), c).ToString(); err != nil || v != "INFO" {
-			t.Fatalf("unexpected response %v %v", v, err)
-		}
-	})
-
 	t.Run("Do", func(t *testing.T) {
 		c := client.B().Get().Key("Do").Build()
 		if v, err := client.Do(context.Background(), c).ToString(); err != nil || v != "GET Do" {
@@ -5090,55 +5066,6 @@ func TestClusterTopologyRefreshment(t *testing.T) {
 			}
 		}
 	})
-
-	t.Run("node role are changed", func(t *testing.T) {
-		var callCount int64
-		refreshWaitCh := make(chan struct{})
-		cli, err := newClusterClient(
-			&ClientOption{
-				InitAddress: []string{"127.0.0.1:0"},
-				ClusterOption: ClusterOption{
-					ShardsRefreshInterval: time.Second,
-				},
-			},
-			func(dst string, opt *ClientOption) conn {
-				return &mockConn{
-					DoFn: func(cmd Completed) RedisResult {
-						if c := atomic.AddInt64(&callCount, 1); c >= 6 {
-							defer func() { recover() }()
-							defer close(refreshWaitCh)
-							return slotsRespWithChangedRole
-						} else if c >= 3 {
-							return slotsRespWithChangedRole
-						}
-						return slotsResp
-					},
-				}
-			},
-			newRetryer(defaultRetryDelayFn),
-		)
-		if err != nil {
-			t.Fatalf("unexpected err %v", err)
-		}
-
-		select {
-		case <-refreshWaitCh:
-			cli.Close()
-
-			cli.mu.Lock()
-			conns := cli.conns
-			cli.mu.Unlock()
-			if len(conns) != 2 {
-				t.Fatalf("unexpected conns %v", conns)
-			}
-			if cc, ok := conns["127.0.0.1:0"]; !ok || !cc.replica {
-				t.Fatalf("unexpected conns %v", conns)
-			}
-			if cc, ok := conns["127.0.1.1:1"]; !ok || cc.replica {
-				t.Fatalf("unexpected conns %v", conns)
-			}
-		}
-	})
 }
 
 func TestClusterClientLoadingRetry(t *testing.T) {