Fix mysql sink may be blocked by network io wait

cdc/sink/mysql.go

@@ -420,10 +420,22 @@ func (s *mysqlSink) createSinkWorkers(ctx context.Context) {
 	}
 }
 
-func (s *mysqlSink) notifyAndWaitExec() {
+func (s *mysqlSink) notifyAndWaitExec(ctx context.Context) {
 	s.notifier.Notify()
-	for _, w := range s.workers {
-		w.waitAllTxnsExecuted()
+	done := make(chan struct{}, 1)
+	go func() {
+		for _, w := range s.workers {
+			w.waitAllTxnsExecuted()
+		}
+		done <- struct{}{}
+	}()
+	// This is a hack code to avoid io wait in some routine blocks others to exit.
+	// As the network io wait is blocked in kernel code, the goroutine is in a
+	// D-state that we could not even stop it by cancel the context. So if this
+	// scenario happens, the blocked goroutine will be leak.
+	select {
+	case <-ctx.Done():
+	case <-done:
 	}
 }
 
@@ -463,7 +475,7 @@ func (s *mysqlSink) dispatchAndExecTxns(ctx context.Context, txnsGroup map[model
 				sendFn(txn, idx)
 				return
 			}
-			s.notifyAndWaitExec()
+			s.notifyAndWaitExec(ctx)
 			causality.reset()
 		}
 		sendFn(txn, rowsChIdx)
@@ -477,7 +489,7 @@ func (s *mysqlSink) dispatchAndExecTxns(ctx context.Context, txnsGroup map[model
 			s.metricConflictDetectDurationHis.Observe(time.Since(startTime).Seconds())
 		}
 	}
-	s.notifyAndWaitExec()
+	s.notifyAndWaitExec(ctx)
 }
 
 type mysqlSinkWorker struct {
@@ -609,6 +621,9 @@ func (s *mysqlSink) execDMLWithMaxRetries(
 			failpoint.Inject("MySQLSinkTxnRandomError", func() {
 				failpoint.Return(checkTxnErr(errors.Trace(dmysql.ErrInvalidConn)))
 			})
+			failpoint.Inject("MySQLSinkHangLongTime", func() {
+				time.Sleep(time.Hour)
+			})
 			err := s.statistics.RecordBatchExecution(func() (int, error) {
 				tx, err := s.db.BeginTx(ctx, nil)
 				if err != nil {
@@ -691,6 +706,9 @@ func (s *mysqlSink) prepareDMLs(rows []*model.RowChangedEvent, replicaID uint64,
 }
 
 func (s *mysqlSink) execDMLs(ctx context.Context, rows []*model.RowChangedEvent, replicaID uint64, bucket int) error {
+	failpoint.Inject("MySQLSinkExecDMLError", func() {
+		failpoint.Return(errors.Trace(dmysql.ErrInvalidConn))
+	})
 	dmls, err := s.prepareDMLs(rows, replicaID, bucket)
 	if err != nil {
 		return errors.Trace(err)

tests/sink_hang/conf/diff_config.toml

@@ -0,0 +1,27 @@
+# diff Configuration.
+
+log-level = "info"
+chunk-size = 10
+check-thread-count = 4
+sample-percent = 100
+use-rowid = false
+use-checksum = true
+fix-sql-file = "fix.sql"
+
+# tables need to check.
+[[check-tables]]
+    schema = "sink_hang"
+    tables = ["~t.*"]
+
+[[source-db]]
+    host = "127.0.0.1"
+    port = 4000
+    user = "root"
+    password = ""
+    instance-id = "source-1"
+
+[target-db]
+    host = "127.0.0.1"
+    port = 3306
+    user = "root"
+    password = ""

tests/sink_hang/run.sh

@@ -0,0 +1,65 @@
+#!/bin/bash
+
+set -e
+
+CUR=$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )
+source $CUR/../_utils/test_prepare
+WORK_DIR=$OUT_DIR/$TEST_NAME
+CDC_BINARY=cdc.test
+SINK_TYPE=$1
+
+CDC_COUNT=3
+DB_COUNT=4
+MAX_RETRIES=10
+
+function check_changefeed_state() {
+    pd_addr=$1
+    changefeed_id=$2
+    expected=$3
+    state=$(cdc cli --pd=$pd_addr changefeed query -s -c $changefeed_id|jq -r ".state")
+    if [[ "$state" != "$expected" ]];then
+        echo "unexpected state $state, expected $expected"
+        exit 1
+    fi
+}
+
+export -f check_changefeed_state
+
+function run() {
+    rm -rf $WORK_DIR && mkdir -p $WORK_DIR
+    start_tidb_cluster --workdir $WORK_DIR
+    cd $WORK_DIR
+
+    pd_addr="http://$UP_PD_HOST:$UP_PD_PORT"
+    TOPIC_NAME="ticdc-sink-hang-test-$RANDOM"
+    case $SINK_TYPE in
+        kafka) SINK_URI="kafka://127.0.0.1:9092/$TOPIC_NAME?partition-num=4";;
+        *) SINK_URI="mysql://[email protected]:3306/?max-txn-row=1";;
+    esac
+    if [ "$SINK_TYPE" == "kafka" ]; then
+      run_kafka_consumer $WORK_DIR "kafka://127.0.0.1:9092/$TOPIC_NAME?partition-num=4"
+    fi
+
+    export GO_FAILPOINTS='github.com/pingcap/ticdc/cdc/sink/MySQLSinkHangLongTime=1*return(true);github.com/pingcap/ticdc/cdc/sink/MySQLSinkExecDMLError=1*return(true)'
+    run_cdc_server --workdir $WORK_DIR --binary $CDC_BINARY --addr "127.0.0.1:8300" --pd $pd_addr
+    changefeed_id=$(cdc cli changefeed create --pd=$pd_addr --sink-uri="$SINK_URI" 2>&1|tail -n2|head -n1|awk '{print $2}')
+
+    run_sql "CREATE DATABASE sink_hang;" ${UP_TIDB_HOST} ${UP_TIDB_PORT}
+    run_sql "CREATE table sink_hang.t1(id int primary key auto_increment, val int);" ${UP_TIDB_HOST} ${UP_TIDB_PORT}
+    run_sql "CREATE table sink_hang.t2(id int primary key auto_increment, val int);" ${UP_TIDB_HOST} ${UP_TIDB_PORT}
+    run_sql "BEGIN; INSERT INTO sink_hang.t1 VALUES (),(),(); INSERT INTO sink_hang.t2 VALUES (),(),(); COMMIT" ${UP_TIDB_HOST} ${UP_TIDB_PORT}
+
+    ensure $MAX_RETRIES check_changefeed_state $pd_addr $changefeed_id "stopped"
+    cdc cli changefeed resume --changefeed-id=$changefeed_id --pd=$pd_addr
+    ensure $MAX_RETRIES check_changefeed_state $pd_addr $changefeed_id "normal"
+
+    check_table_exists "sink_hang.t1" ${DOWN_TIDB_HOST} ${DOWN_TIDB_PORT}
+    check_table_exists "sink_hang.t2" ${DOWN_TIDB_HOST} ${DOWN_TIDB_PORT}
+    check_sync_diff $WORK_DIR $CUR/conf/diff_config.toml
+
+    cleanup_process $CDC_BINARY
+}
+
+trap stop_tidb_cluster EXIT
+run $*
+echo "[$(date)] <<<<<< run test case $TEST_NAME success! >>>>>>"