metrics: remove metrics proxy (#6948)

ref #4287

dm/dumpling/dumpling.go

@@ -30,7 +30,6 @@ import (
 	"go.uber.org/atomic"
 	"go.uber.org/zap"
 
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 	"github.com/pingcap/tiflow/engine/pkg/promutil"
 
 	"github.com/pingcap/tiflow/dm/config"
@@ -81,8 +80,7 @@ func (m *Dumpling) Init(ctx context.Context) error {
 		// will register and deregister metrics, so we must use NoopRegistry
 		// to avoid duplicated registration.
 		m.metricProxies = &metricProxies{}
-		m.metricProxies.dumplingExitWithErrorCounter = metricsproxy.NewCounterVec(
-			m.cfg.MetricsFactory,
+		m.metricProxies.dumplingExitWithErrorCounter = m.cfg.MetricsFactory.NewCounterVec(
 			prometheus.CounterOpts{
 				Namespace: "dm",
 				Subsystem: "dumpling",

dm/dumpling/metrics.go

@@ -14,18 +14,18 @@
 package dumpling
 
 import (
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 	"github.com/pingcap/tiflow/engine/pkg/promutil"
 	"github.com/prometheus/client_golang/prometheus"
 )
 
 type metricProxies struct {
-	dumplingExitWithErrorCounter *metricsproxy.CounterVecProxy
+	dumplingExitWithErrorCounter *prometheus.CounterVec
 }
 
+var f = &promutil.PromFactory{}
+
 var defaultMetricProxies = &metricProxies{
-	dumplingExitWithErrorCounter: metricsproxy.NewCounterVec(
-		&promutil.PromFactory{},
+	dumplingExitWithErrorCounter: f.NewCounterVec(
 		prometheus.CounterOpts{
 			Namespace: "dm",
 			Subsystem: "dumpling",
@@ -41,5 +41,5 @@ func RegisterMetrics(registry *prometheus.Registry) {
 
 func (m *Dumpling) removeLabelValuesWithTaskInMetrics(task, source string) {
 	labels := prometheus.Labels{"task": task, "source_id": source}
-	m.metricProxies.dumplingExitWithErrorCounter.DeleteAllAboutLabels(labels)
+	m.metricProxies.dumplingExitWithErrorCounter.DeletePartialMatch(labels)
 }

dm/loader/metrics.go

@@ -16,21 +16,20 @@ package loader
 import (
 	"github.com/pingcap/tiflow/engine/pkg/promutil"
 	"github.com/prometheus/client_golang/prometheus"
-
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 )
 
 var (
+	f = &promutil.PromFactory{}
 	// should error.
-	tidbExecutionErrorCounter = metricsproxy.NewCounterVec(&promutil.PromFactory{},
+	tidbExecutionErrorCounter = f.NewCounterVec(
 		prometheus.CounterOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
 			Name:      "tidb_execution_error",
 			Help:      "Total count of tidb execution errors",
 		}, []string{"task", "source_id"})
 
-	queryHistogram = metricsproxy.NewHistogramVec(&promutil.PromFactory{},
+	queryHistogram = f.NewHistogramVec(
 		prometheus.HistogramOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
@@ -39,7 +38,7 @@ var (
 			Buckets:   prometheus.ExponentialBuckets(0.000005, 2, 25),
 		}, []string{"task", "source_id"})
 
-	txnHistogram = metricsproxy.NewHistogramVec(&promutil.PromFactory{},
+	txnHistogram = f.NewHistogramVec(
 		prometheus.HistogramOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
@@ -48,7 +47,7 @@ var (
 			Buckets:   prometheus.ExponentialBuckets(0.000005, 2, 25),
 		}, []string{"task", "worker", "source_id", "target_schema", "target_table"})
 
-	stmtHistogram = metricsproxy.NewHistogramVec(&promutil.PromFactory{},
+	stmtHistogram = f.NewHistogramVec(
 		prometheus.HistogramOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
@@ -57,31 +56,31 @@ var (
 			Buckets:   prometheus.ExponentialBuckets(0.000005, 2, 25),
 		}, []string{"type", "task"})
 
-	dataFileGauge = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	dataFileGauge = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
 			Name:      "data_file_gauge",
 			Help:      "data files in total",
 		}, []string{"task", "source_id"})
 
-	tableGauge = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	tableGauge = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
 			Name:      "table_gauge",
 			Help:      "tables in total",
 		}, []string{"task", "source_id"})
 
-	dataSizeGauge = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	dataSizeGauge = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
 			Name:      "data_size_gauge",
 			Help:      "data size in total",
 		}, []string{"task", "source_id"})
 
-	progressGauge = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	progressGauge = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
@@ -90,15 +89,15 @@ var (
 		}, []string{"task", "source_id"})
 
 	// should alert.
-	loaderExitWithErrorCounter = metricsproxy.NewCounterVec(&promutil.PromFactory{},
+	loaderExitWithErrorCounter = f.NewCounterVec(
 		prometheus.CounterOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
 			Name:      "exit_with_error_count",
 			Help:      "counter for loader exits with error",
 		}, []string{"task", "source_id"})
 
-	remainingTimeGauge = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	remainingTimeGauge = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "loader",
@@ -122,14 +121,14 @@ func RegisterMetrics(registry *prometheus.Registry) {
 }
 
 func (l *Loader) removeLabelValuesWithTaskInMetrics(task string) {
-	tidbExecutionErrorCounter.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	txnHistogram.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	queryHistogram.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	stmtHistogram.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	dataFileGauge.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	tableGauge.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	dataSizeGauge.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	progressGauge.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	loaderExitWithErrorCounter.DeleteAllAboutLabels(prometheus.Labels{"task": task})
-	remainingTimeGauge.DeleteAllAboutLabels(prometheus.Labels{"task": task})
+	tidbExecutionErrorCounter.DeletePartialMatch(prometheus.Labels{"task": task})
+	txnHistogram.DeletePartialMatch(prometheus.Labels{"task": task})
+	queryHistogram.DeletePartialMatch(prometheus.Labels{"task": task})
+	stmtHistogram.DeletePartialMatch(prometheus.Labels{"task": task})
+	dataFileGauge.DeletePartialMatch(prometheus.Labels{"task": task})
+	tableGauge.DeletePartialMatch(prometheus.Labels{"task": task})
+	dataSizeGauge.DeletePartialMatch(prometheus.Labels{"task": task})
+	progressGauge.DeletePartialMatch(prometheus.Labels{"task": task})
+	loaderExitWithErrorCounter.DeletePartialMatch(prometheus.Labels{"task": task})
+	remainingTimeGauge.DeletePartialMatch(prometheus.Labels{"task": task})
 }

dm/master/metrics/metrics.go

@@ -20,8 +20,6 @@ import (
 	cpu "github.com/pingcap/tidb-tools/pkg/utils"
 	"github.com/pingcap/tiflow/engine/pkg/promutil"
 	"github.com/prometheus/client_golang/prometheus"
-
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 )
 
 // used for ddlPendingCounter, no "Resolved" lock because they will be
@@ -48,7 +46,8 @@ const (
 )
 
 var (
-	workerState = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	f           = &promutil.PromFactory{}
+	workerState = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "master",
@@ -64,23 +63,23 @@ var (
 			Help:      "the cpu usage of master",
 		})
 
-	ddlPendingCounter = metricsproxy.NewGaugeVec(&promutil.PromFactory{},
+	ddlPendingCounter = f.NewGaugeVec(
 		prometheus.GaugeOpts{
 			Namespace: "dm",
 			Subsystem: "master",
 			Name:      "ddl_state_number",
 			Help:      "number of pending DDL in different states, Un-synced (waiting all upstream), Synced (all upstream finished, waiting all downstream)",
 		}, []string{"task", "type"})
 
-	ddlErrCounter = metricsproxy.NewCounterVec(&promutil.PromFactory{},
+	ddlErrCounter = f.NewCounterVec(
 		prometheus.CounterOpts{
 			Namespace: "dm",
 			Subsystem: "master",
 			Name:      "shard_ddl_error",
 			Help:      "number of shard DDL lock/operation error",
 		}, []string{"task", "type"})
 
-	workerEventErrCounter = metricsproxy.NewCounterVec(&promutil.PromFactory{},
+	workerEventErrCounter = f.NewCounterVec(
 		prometheus.CounterOpts{
 			Namespace: "dm",
 			Subsystem: "master",
@@ -137,7 +136,7 @@ func ReportWorkerStage(name string, state float64) {
 
 // RemoveWorkerState cleans state of deleted worker.
 func RemoveWorkerState(name string) {
-	workerState.DeleteAllAboutLabels(prometheus.Labels{"worker": name})
+	workerState.DeletePartialMatch(prometheus.Labels{"worker": name})
 }
 
 // ReportDDLPending inc/dec by 1 to ddlPendingCounter.
@@ -152,7 +151,7 @@ func ReportDDLPending(task, oldStatus, newStatus string) {
 
 // RemoveDDLPending removes all counter of this task.
 func RemoveDDLPending(task string) {
-	ddlPendingCounter.DeleteAllAboutLabels(prometheus.Labels{"task": task})
+	ddlPendingCounter.DeletePartialMatch(prometheus.Labels{"task": task})
 }
 
 // ReportDDLError is a setter for ddlErrCounter.

dm/pkg/conn/baseconn.go

@@ -23,11 +23,11 @@ import (
 	gmysql "github.com/go-mysql-org/go-mysql/mysql"
 	"github.com/go-sql-driver/mysql"
 	"github.com/pingcap/failpoint"
+	"github.com/prometheus/client_golang/prometheus"
 	"go.uber.org/zap"
 
 	tcontext "github.com/pingcap/tiflow/dm/pkg/context"
 	"github.com/pingcap/tiflow/dm/pkg/log"
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 	"github.com/pingcap/tiflow/dm/pkg/retry"
 	"github.com/pingcap/tiflow/dm/pkg/terror"
 	"github.com/pingcap/tiflow/dm/pkg/utils"
@@ -117,7 +117,7 @@ func (conn *BaseConn) QuerySQL(tctx *tcontext.Context, query string, args ...int
 // return
 // 1. failed: (the index of sqls executed error, error)
 // 2. succeed: (len(sqls), nil).
-func (conn *BaseConn) ExecuteSQLWithIgnoreError(tctx *tcontext.Context, hVec *metricsproxy.HistogramVecProxy, task string, ignoreErr func(error) bool, queries []string, args ...[]interface{}) (int, error) {
+func (conn *BaseConn) ExecuteSQLWithIgnoreError(tctx *tcontext.Context, hVec *prometheus.HistogramVec, task string, ignoreErr func(error) bool, queries []string, args ...[]interface{}) (int, error) {
 	// inject an error to trigger retry, this should be placed before the real execution of the SQL statement.
 	failpoint.Inject("retryableError", func(val failpoint.Value) {
 		if mark, ok := val.(string); ok {
@@ -216,7 +216,7 @@ func (conn *BaseConn) ExecuteSQLWithIgnoreError(tctx *tcontext.Context, hVec *me
 // return
 // 1. failed: (the index of sqls executed error, error)
 // 2. succeed: (len(sqls), nil).
-func (conn *BaseConn) ExecuteSQL(tctx *tcontext.Context, hVec *metricsproxy.HistogramVecProxy, task string, queries []string, args ...[]interface{}) (int, error) {
+func (conn *BaseConn) ExecuteSQL(tctx *tcontext.Context, hVec *prometheus.HistogramVec, task string, queries []string, args ...[]interface{}) (int, error) {
 	return conn.ExecuteSQLWithIgnoreError(tctx, hVec, task, nil, queries, args...)
 }
 

dm/pkg/conn/baseconn_test.go

@@ -19,7 +19,6 @@ import (
 	"testing"
 
 	tcontext "github.com/pingcap/tiflow/dm/pkg/context"
-	"github.com/pingcap/tiflow/dm/pkg/metricsproxy"
 	"github.com/pingcap/tiflow/dm/pkg/retry"
 	"github.com/pingcap/tiflow/dm/pkg/terror"
 	"github.com/pingcap/tiflow/engine/pkg/promutil"
@@ -37,7 +36,9 @@ var _ = Suite(&testBaseConnSuite{})
 
 type testBaseConnSuite struct{}
 
-var testStmtHistogram = metricsproxy.NewHistogramVec(&promutil.PromFactory{},
+var f = &promutil.PromFactory{}
+
+var testStmtHistogram = f.NewHistogramVec(
 	prometheus.HistogramOpts{
 		Namespace: "dm",
 		Subsystem: "conn",