*: reset config if the input is invalid (#8632) (#8659)

close #8619 Signed-off-by: ti-chi-bot <[email protected]> Signed-off-by: Ryan Leung <[email protected]> Co-authored-by: Ryan Leung <[email protected]>
tikv · Sep 24, 2024 · f2231c2 · f2231c2
1 parent eeead6b
commit f2231c2
Show file tree

Hide file tree

Showing 4 changed files with 94 additions and 6 deletions.
diff --git a/pkg/schedule/schedulers/evict_leader.go b/pkg/schedule/schedulers/evict_leader.go
@@ -74,6 +74,9 @@ func (conf *evictLeaderSchedulerConfig) getStores() []uint64 {
 }
 
 func (conf *evictLeaderSchedulerConfig) BuildWithArgs(args []string) error {
+	failpoint.Inject("buildWithArgsErr", func() {
+		failpoint.Return(errors.New("fail to build with args"))
+	})
 	if len(args) != 1 {
 		return errs.ErrSchedulerConfig.FastGenByArgs("id")
 	}
@@ -401,8 +404,15 @@ func (handler *evictLeaderHandler) UpdateConfig(w http.ResponseWriter, r *http.R
 		args = append(args, handler.config.getRanges(id)...)
 	}
 
-	handler.config.BuildWithArgs(args)
-	err := handler.config.Persist()
+	err := handler.config.BuildWithArgs(args)
+	if err != nil {
+		handler.config.mu.Lock()
+		handler.config.cluster.ResumeLeaderTransfer(id)
+		handler.config.mu.Unlock()
+		handler.rd.JSON(w, http.StatusBadRequest, err.Error())
+		return
+	}
+	err = handler.config.Persist()
 	if err != nil {
 		handler.config.removeStore(id)
 		handler.rd.JSON(w, http.StatusInternalServerError, err.Error())

diff --git a/pkg/schedule/schedulers/grant_leader.go b/pkg/schedule/schedulers/grant_leader.go
@@ -287,8 +287,15 @@ func (handler *grantLeaderHandler) UpdateConfig(w http.ResponseWriter, r *http.R
 		args = append(args, handler.config.getRanges(id)...)
 	}
 
-	handler.config.BuildWithArgs(args)
-	err := handler.config.Persist()
+	err := handler.config.BuildWithArgs(args)
+	if err != nil {
+		handler.config.mu.Lock()
+		handler.config.cluster.ResumeLeaderTransfer(id)
+		handler.config.mu.Unlock()
+		handler.rd.JSON(w, http.StatusBadRequest, err.Error())
+		return
+	}
+	err = handler.config.Persist()
 	if err != nil {
 		handler.config.removeStore(id)
 		handler.rd.JSON(w, http.StatusInternalServerError, err.Error())

diff --git a/plugin/scheduler_example/evict_leader.go b/plugin/scheduler_example/evict_leader.go
@@ -275,9 +275,21 @@ func (handler *evictLeaderHandler) UpdateConfig(w http.ResponseWriter, r *http.R
 		args = append(args, handler.config.getRanges(id)...)
 	}
 
-	handler.config.BuildWithArgs(args)
-	err := handler.config.Persist()
+	err := handler.config.BuildWithArgs(args)
 	if err != nil {
+		handler.config.mu.Lock()
+		handler.config.cluster.ResumeLeaderTransfer(id)
+		handler.config.mu.Unlock()
+		handler.rd.JSON(w, http.StatusBadRequest, err.Error())
+		return
+	}
+
+	err = handler.config.Persist()
+	if err != nil {
+		handler.config.mu.Lock()
+		delete(handler.config.StoreIDWitRanges, id)
+		handler.config.cluster.ResumeLeaderTransfer(id)
+		handler.config.mu.Unlock()
 		handler.rd.JSON(w, http.StatusInternalServerError, err.Error())
 	}
 	handler.rd.JSON(w, http.StatusOK, nil)

diff --git a/tests/pdctl/scheduler/scheduler_test.go b/tests/pdctl/scheduler/scheduler_test.go
@@ -20,6 +20,7 @@ import (
 	"testing"
 	"time"
 
+	"github.com/pingcap/failpoint"
 	"github.com/pingcap/kvproto/pkg/metapb"
 	"github.com/spf13/cobra"
 	"github.com/stretchr/testify/require"
@@ -584,3 +585,61 @@ func TestForwardSchedulerRequest(t *testing.T) {
 		"balance-leader-scheduler",
 	})
 }
+
+func TestEvictLeaderScheduler(t *testing.T) {
+	re := require.New(t)
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+	cluster, err := tests.NewTestCluster(ctx, 1)
+	re.NoError(err)
+	defer cluster.Destroy()
+	err = cluster.RunInitialServers()
+	re.NoError(err)
+	cluster.WaitLeader()
+	pdAddr := cluster.GetConfig().GetClientURL()
+	cmd := pdctlCmd.GetRootCmd()
+
+	stores := []*metapb.Store{
+		{
+			Id:            1,
+			State:         metapb.StoreState_Up,
+			LastHeartbeat: time.Now().UnixNano(),
+		},
+		{
+			Id:            2,
+			State:         metapb.StoreState_Up,
+			LastHeartbeat: time.Now().UnixNano(),
+		},
+		{
+			Id:            3,
+			State:         metapb.StoreState_Up,
+			LastHeartbeat: time.Now().UnixNano(),
+		},
+		{
+			Id:            4,
+			State:         metapb.StoreState_Up,
+			LastHeartbeat: time.Now().UnixNano(),
+		},
+	}
+	leaderServer := cluster.GetServer(cluster.GetLeader())
+	re.NoError(leaderServer.BootstrapCluster())
+	for _, store := range stores {
+		tests.MustPutStore(re, cluster, store)
+	}
+
+	tests.MustPutRegion(re, cluster, 1, 1, []byte("a"), []byte("b"))
+	output, err := pdctl.ExecuteCommand(cmd, []string{"-u", pdAddr, "scheduler", "add", "evict-leader-scheduler", "2"}...)
+	re.NoError(err)
+	re.Contains(string(output), "Success!")
+	failpoint.Enable("github.com/tikv/pd/pkg/schedule/schedulers/buildWithArgsErr", "return(true)")
+	output, err = pdctl.ExecuteCommand(cmd, []string{"-u", pdAddr, "scheduler", "add", "evict-leader-scheduler", "1"}...)
+	re.NoError(err)
+	re.Contains(string(output), "fail to build with args")
+	failpoint.Disable("github.com/tikv/pd/pkg/schedule/schedulers/buildWithArgsErr")
+	output, err = pdctl.ExecuteCommand(cmd, []string{"-u", pdAddr, "scheduler", "remove", "evict-leader-scheduler"}...)
+	re.NoError(err)
+	re.Contains(string(output), "Success!")
+	output, err = pdctl.ExecuteCommand(cmd, []string{"-u", pdAddr, "scheduler", "add", "evict-leader-scheduler", "1"}...)
+	re.NoError(err)
+	re.Contains(string(output), "Success!")
+}