Merge remote-tracking branch 'upstream/master' into add_store_label

.github/workflows/pd-tests.yaml

@@ -31,15 +31,23 @@ jobs:
           - worker_id: 2
             name: 'Unit Test(2)'
           - worker_id: 3
-            name: 'Tools Test'
+            name: 'Unit Test(3)'
           - worker_id: 4
-            name: 'Client Integration Test'
+            name: 'Tests(1)'
           - worker_id: 5
-            name: 'TSO Integration Test'
+            name: 'Tests(2)'
           - worker_id: 6
-            name: 'MicroService Integration Test'
+            name: 'Tools Test'
+          - worker_id: 7
+            name: 'Client Integration Test'
+          - worker_id: 8
+            name: 'TSO Integration Test'
+          - worker_id: 9
+            name: 'MicroService Integration(!TSO)'
+          - worker_id: 10
+            name: 'MicroService Integration(TSO)'
     outputs:
-      job-total: 6
+      job-total: 10
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
@@ -52,6 +60,9 @@ jobs:
         run: |
           make ci-test-job JOB_INDEX=$WORKER_ID
           mv covprofile covprofile_$WORKER_ID
+          if [ -f junitfile ]; then
+            cat junitfile
+          fi
       - name: Upload coverage result ${{ matrix.worker_id }}
         uses: actions/upload-artifact@v4
         with:
@@ -77,7 +88,7 @@ jobs:
           # only keep the first line(`mode: aomic`) of the coverage profile
           sed -i '2,${/mode: atomic/d;}' covprofile
       - name: Send coverage
-        uses: codecov/codecov-action@v4.2.0
+        uses: codecov/codecov-action@v4
         with:
           token: ${{ secrets.CODECOV }}
           file: ./covprofile

.gitignore

@@ -19,6 +19,7 @@ docs/swagger/*
 *.before
 covprofile
 coverage.tmp
+junitfile
 package.list
 report.xml
 coverage.xml

.golangci.yml

@@ -198,6 +198,14 @@ linters-settings:
         severity: warning
         disabled: false
         exclude: [""]
+  errcheck:
+    exclude-functions:
+      - (*github.com/unrolled/render.Render).JSON
+      - (*github.com/unrolled/render.Render).Data
+      - (*github.com/unrolled/render.Render).Text
+      - (net/http.ResponseWriter).Write
+      - github.com/pingcap/log.Sync
+      - (github.com/tikv/pd/pkg/ratelimit.Runner).RunTask
 issues:
   exclude-rules:
     - path: (_test\.go|pkg/mock/.*\.go|tests/.*\.go)
@@ -207,6 +215,3 @@ issues:
     - path: (pd-analysis|pd-api-bench|pd-backup|pd-ctl|pd-heartbeat-bench|pd-recover|pd-simulator|pd-tso-bench|pd-ut|regions-dump|stores-dump)
       linters:
         - errcheck
-    - path: (pkg/tso/admin.go|pkg/schedule/schedulers/split_bucket.go|server/api/plugin_disable.go|server/api/plugin_disable.go|server/api/operator.go|server/api/region.go|pkg/schedule/schedulers/balance_leader.go|pkg/replication/replication_mode.go|pkg/storage/endpoint/gc_safe_point.go|server/.*\.go|pkg/syncer/server.go)
-      linters:
-        - errcheck

Makefile

@@ -228,7 +228,7 @@ failpoint-disable: install-tools
 ut: pd-ut
 	@$(FAILPOINT_ENABLE)
 	# only run unit tests
-	./bin/pd-ut run --ignore tests --race
+	./bin/pd-ut run --ignore tests --race --junitfile ./junitfile
 	@$(CLEAN_UT_BINARY)
 	@$(FAILPOINT_DISABLE)
 
@@ -254,7 +254,7 @@ basic-test: install-tools
 
 ci-test-job: install-tools dashboard-ui pd-ut
 	@$(FAILPOINT_ENABLE)
-	./scripts/ci-subtask.sh $(JOB_COUNT) $(JOB_INDEX) || { $(FAILPOINT_DISABLE); exit 1; }
+	./scripts/ci-subtask.sh $(JOB_INDEX) || { $(FAILPOINT_DISABLE); exit 1; }
 	@$(FAILPOINT_DISABLE)
 
 TSO_INTEGRATION_TEST_PKGS := $(PD_PKG)/tests/server/tso

client/go.mod

@@ -33,6 +33,7 @@ require (
 	github.com/prometheus/client_model v0.5.0 // indirect
 	github.com/prometheus/common v0.46.0 // indirect
 	github.com/prometheus/procfs v0.12.0 // indirect
+	github.com/stretchr/objx v0.5.0 // indirect
 	go.uber.org/multierr v1.11.0 // indirect
 	golang.org/x/net v0.23.0 // indirect
 	golang.org/x/sys v0.18.0 // indirect

client/go.sum

@@ -68,6 +68,7 @@ github.com/rogpeppe/go-internal v1.10.0/go.mod h1:UQnix2H7Ngw/k4C5ijL5+65zddjncj
 github.com/sergi/go-diff v1.1.0/go.mod h1:STckp+ISIX8hZLjrqAeVduY0gWCT9IjLuqbuNXdaHfM=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/objx v0.4.0/go.mod h1:YvHI0jy2hoMjB+UWwv71VJQ9isScKT/TqJzVSSt89Yw=
+github.com/stretchr/objx v0.5.0 h1:1zr/of2m5FGMsad5YfcqgdqdWrIhu+EBEJRhR1U7z/c=
 github.com/stretchr/objx v0.5.0/go.mod h1:Yh+to48EsGEfYuaHDzXPcE3xhTkx73EhmCGUpEOglKo=
 github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=
 github.com/stretchr/testify v1.4.0/go.mod h1:j7eGeouHqKxXV5pUuKE4zz7dFj8WfuZ+81PSLYec5m4=

client/metrics.go

@@ -105,7 +105,7 @@ func initMetrics(constLabels prometheus.Labels) {
 			Subsystem:   "request",
 			Name:        "tso_batch_send_latency",
 			ConstLabels: constLabels,
-			Buckets:     prometheus.ExponentialBuckets(1, 2, 34), // 1ns ~ 8s
+			Buckets:     prometheus.ExponentialBuckets(0.0005, 2, 13),
 			Help:        "tso batch send latency",
 		})
 

client/resource_group/controller/config.go

@@ -52,8 +52,10 @@ const (
 	defaultTargetPeriod = 5 * time.Second
 	// defaultMaxWaitDuration is the max duration to wait for the token before throwing error.
 	defaultMaxWaitDuration = 30 * time.Second
+	// defaultLTBTokenRPCMaxDelay is the upper bound of backoff delay for local token bucket RPC.
+	defaultLTBTokenRPCMaxDelay = 1 * time.Second
 	// defaultWaitRetryTimes is the times to retry when waiting for the token.
-	defaultWaitRetryTimes = 10
+	defaultWaitRetryTimes = 20
 	// defaultWaitRetryInterval is the interval to retry when waiting for the token.
 	defaultWaitRetryInterval = 50 * time.Millisecond
 )
@@ -77,23 +79,35 @@ const (
 
 	// Because the resource manager has not been deployed in microservice mode,
 	// do not enable this function.
-	defaultDegradedModeWaitDuration = 0
+	defaultDegradedModeWaitDuration = time.Duration(0)
 	defaultAvgBatchProportion       = 0.7
 )
 
-// Config is the configuration of the resource manager controller which includes some option for client needed.
-type Config struct {
+// TokenRPCParams is the parameters for local bucket RPC.
+type TokenRPCParams struct {
+	// WaitRetryInterval is the interval to retry when waiting for the token.
+	WaitRetryInterval Duration `toml:"wait-retry-interval" json:"wait-retry-interval"`
+
+	// WaitRetryTimes is the times to retry when waiting for the token.
+	WaitRetryTimes int `toml:"wait-retry-times" json:"wait-retry-times"`
+}
+
+// LocalBucketConfig is the configuration for local bucket. not export to server side.
+type LocalBucketConfig struct {
+	TokenRPCParams `toml:"token-rpc-params" json:"token-rpc-params"`
+}
+
+// BaseConfig is the configuration of the resource manager controller which includes some option for client needed.
+// TODO: unified the configuration for client and server, server side in pkg/mcs/resourcemanger/config.go.
+type BaseConfig struct {
 	// EnableDegradedMode is to control whether resource control client enable degraded mode when server is disconnect.
 	DegradedModeWaitDuration Duration `toml:"degraded-mode-wait-duration" json:"degraded-mode-wait-duration"`
 
 	// LTBMaxWaitDuration is the max wait time duration for local token bucket.
 	LTBMaxWaitDuration Duration `toml:"ltb-max-wait-duration" json:"ltb-max-wait-duration"`
 
-	// WaitRetryInterval is the interval to retry when waiting for the token.
-	WaitRetryInterval Duration `toml:"wait-retry-interval" json:"wait-retry-interval"`
-
-	// WaitRetryTimes is the times to retry when waiting for the token.
-	WaitRetryTimes int `toml:"wait-retry-times" json:"wait-retry-times"`
+	// LTBTokenRPCMaxDelay is the upper bound of backoff delay for local token bucket RPC.
+	LTBTokenRPCMaxDelay Duration `toml:"ltb-token-rpc-max-delay" json:"ltb-token-rpc-max-delay"`
 
 	// RequestUnit is the configuration determines the coefficients of the RRU and WRU cost.
 	// This configuration should be modified carefully.
@@ -103,15 +117,43 @@ type Config struct {
 	EnableControllerTraceLog bool `toml:"enable-controller-trace-log" json:"enable-controller-trace-log,string"`
 }
 
+// Config is the configuration of the resource manager controller.
+type Config struct {
+	BaseConfig
+	LocalBucketConfig
+}
+
+// Adjust adjusts the configuration.
+func (c *Config) Adjust() {
+	// valid the configuration, TODO: separately add the valid function.
+	if c.BaseConfig.LTBMaxWaitDuration.Duration == 0 {
+		c.BaseConfig.LTBMaxWaitDuration = NewDuration(defaultMaxWaitDuration)
+	}
+	if c.LocalBucketConfig.WaitRetryInterval.Duration == 0 {
+		c.LocalBucketConfig.WaitRetryInterval = NewDuration(defaultWaitRetryInterval)
+	}
+	// adjust the client settings. calculate the retry times.
+	if int(c.BaseConfig.LTBTokenRPCMaxDelay.Duration) != int(c.LocalBucketConfig.WaitRetryInterval.Duration)*c.LocalBucketConfig.WaitRetryTimes {
+		c.LocalBucketConfig.WaitRetryTimes = int(c.BaseConfig.LTBTokenRPCMaxDelay.Duration / c.LocalBucketConfig.WaitRetryInterval.Duration)
+	}
+}
+
 // DefaultConfig returns the default resource manager controller configuration.
 func DefaultConfig() *Config {
 	return &Config{
-		DegradedModeWaitDuration: NewDuration(defaultDegradedModeWaitDuration),
-		LTBMaxWaitDuration:       NewDuration(defaultMaxWaitDuration),
-		WaitRetryInterval:        NewDuration(defaultWaitRetryInterval),
-		WaitRetryTimes:           defaultWaitRetryTimes,
-		RequestUnit:              DefaultRequestUnitConfig(),
-		EnableControllerTraceLog: false,
+		BaseConfig: BaseConfig{
+			DegradedModeWaitDuration: NewDuration(defaultDegradedModeWaitDuration),
+			RequestUnit:              DefaultRequestUnitConfig(),
+			EnableControllerTraceLog: false,
+			LTBMaxWaitDuration:       NewDuration(defaultMaxWaitDuration),
+			LTBTokenRPCMaxDelay:      NewDuration(defaultLTBTokenRPCMaxDelay),
+		},
+		LocalBucketConfig: LocalBucketConfig{
+			TokenRPCParams: TokenRPCParams{
+				WaitRetryInterval: NewDuration(defaultWaitRetryInterval),
+				WaitRetryTimes:    defaultWaitRetryTimes,
+			},
+		},
 	}
 }