TestHTTPClientTestSuiteWithServiceDiscovery is unstable

[lhy1024]

Flaky Test

Which jobs are failing

FAIL: TestHTTPClientTestSuiteWithServiceDiscovery/TestTransferLeader

CI link

https://github.com/tikv/pd/actions/runs/10487662345/job/29048498057

2024-08-21T09:59:37.1850214Z     http_client_test.go:637: 
2024-08-21T09:59:37.1851322Z         	Error Trace:	/home/runner/work/pd/pd/tests/integrations/client/http_client_test.go:637
2024-08-21T09:59:37.1852262Z         	Error:      	Should not be: "pd1"
2024-08-21T09:59:37.1853014Z         	Test:       	TestHTTPClientTestSuiteWithServiceDiscovery/TestTransferLeader

Reason for failure (if possible)

Anything else

[HuSharp]

We can find PD2 campaign times by log:

1. SetupSuite make PD2 as leader ----- 1

2. TestRetryOnLeaderChange

TestRetryOnLeaderChange is before TestTransferLeader, which have 3 times resign

Pd2->PD1
Pd1->PD2 -------- 2
Pd2->PD1 --------- during this resign etcd had stuck which make PD2 still can start campaign leader

This is the reason why the test is unstable, if it's not stuck, for the whole test, here the campaign times won't +1

2024-08-21T09:59:37.1252137Z [2024/08/21 09:59:21.828 +00:00] [INFO] [member.go:362] ["try to resign etcd leader to next pd-server"] [from=pd2] [to=]
2024-08-21T09:59:37.1253432Z [2024/08/21 09:59:21.828 +00:00] [INFO] [cluster.go:710] ["node state check job has been stopped"]
2024-08-21T09:59:37.1254169Z [2024/08/21 09:59:21.828 +00:00] [INFO] [cluster.go:686] ["metrics are reset"]
2024-08-21T09:59:37.1254945Z [2024/08/21 09:59:21.828 +00:00] [INFO] [server.go:127] ["region syncer has been stopped"]
2024-08-21T09:59:37.1255938Z [2024/08/21 09:59:21.828 +00:00] [INFO] [cluster.go:688] ["metrics collection job has been stopped"]
2024-08-21T09:59:37.1256731Z [2024/08/21 09:59:21.828 +00:00] [INFO] [cluster.go:520] ["sync store config job is stopped"]
2024-08-21T09:59:37.1257462Z [2024/08/21 09:59:21.828 +00:00] [INFO] [cluster.go:407] ["service check job is stopped"]
2024-08-21T09:59:37.1258166Z [2024/08/21 09:59:21.828 +00:00] [INFO] [labeler.go:70] ["region labeler GC stopped"]
2024-08-21T09:59:37.1259073Z [2024/08/21 09:59:21.828 +00:00] [INFO] [runner.go:131] ["stopping async task runner"] [name=heartbeat-async]
2024-08-21T09:59:37.1260105Z [2024/08/21 09:59:21.828 +00:00] [INFO] [runner.go:131] ["stopping async task runner"] [name=log-async]
2024-08-21T09:59:37.1261069Z [2024/08/21 09:59:21.828 +00:00] [INFO] [runner.go:131] ["stopping async task runner"] [name=misc-async]
2024-08-21T09:59:37.1261982Z [2024/08/21 09:59:21.828 +00:00] [INFO] [scheduling_controller.go:161] ["scheduling metrics are reset"]
2024-08-21T09:59:37.1263113Z [2024/08/21 09:59:21.828 +00:00] [INFO] [scheduling_controller.go:163] ["scheduling metrics collection job has been stopped"]
2024-08-21T09:59:37.1263986Z [2024/08/21 09:59:21.828 +00:00] [INFO] [coordinator.go:232] ["coordinator stops running"]
2024-08-21T09:59:37.1264715Z [2024/08/21 09:59:21.829 +00:00] [INFO] [coordinator.go:210] ["coordinator is stopping"]
2024-08-21T09:59:37.1265559Z [2024/08/21 09:59:21.829 +00:00] [INFO] [scheduling_controller.go:139] ["statistics background jobs has been stopped"]
2024-08-21T09:59:37.1266409Z [2024/08/21 09:59:21.829 +00:00] [INFO] [coordinator.go:213] ["coordinator has been stopped"]
2024-08-21T09:59:37.1267216Z [2024/08/21 09:59:21.829 +00:00] [INFO] [scheduling_controller.go:87] ["scheduling service is stopped"]
2024-08-21T09:59:37.1267975Z [2024/08/21 09:59:21.830 +00:00] [INFO] [cluster.go:769] ["raft cluster is stopped"]
2024-08-21T09:59:37.1268676Z [2024/08/21 09:59:21.830 +00:00] [INFO] [tso.go:436] ["reset the timestamp in memory"] []
2024-08-21T09:59:37.1270080Z [2024/08/21 09:59:21.831 +00:00] [INFO] [server.go:1481] ["leadership transfer starting"] [local-member-id=e298409efaeb2293] [current-leader-member-id=e298409efaeb2293] [transferee-member-id=319ee8b43d843ed7]
2024-08-21T09:59:37.1272407Z [2024/08/21 09:59:21.831 +00:00] [INFO] [raft] [zap_raft.go:77] ["e298409efaeb2293 [term 4] starts to transfer leadership to 319ee8b43d843ed7"]
2024-08-21T09:59:37.1275064Z [2024/08/21 09:59:21.831 +00:00] [INFO] [raft] [zap_raft.go:77] ["e298409efaeb2293 sends MsgTimeoutNow to 319ee8b43d843ed7 immediately as 319ee8b43d843ed7 already has up-to-date log"]
2024-08-21T09:59:37.1277696Z [2024/08/21 09:59:21.831 +00:00] [INFO] [raft] [zap_raft.go:77] ["319ee8b43d843ed7 [term 4] received MsgTimeoutNow from e298409efaeb2293 and starts an election to get leadership."]
2024-08-21T09:59:37.1279845Z [2024/08/21 09:59:21.831 +00:00] [INFO] [raft] [zap_raft.go:77] ["319ee8b43d843ed7 is starting a new election at term 4"]
2024-08-21T09:59:37.1281814Z [2024/08/21 09:59:21.831 +00:00] [INFO] [server.go:1690] ["start to campaign PD leader"] [campaign-leader-name=pd2]

3. start TestTransferLeader
   PD1 -> PD2 ------- 4 And then meet frequently error