testStatsSuite.TestAnalyzeGlobalStatsWithOpts & testStatsSuite.TestAnalyzeGlobalStatsWithOpts2 failed when make gotest

[purelind]

Bug Report

make gotest
https://ci.pingcap.net/blue/organizations/jenkins/tidb_ghpr_coverage/detail/tidb_ghpr_coverage/138/pipeline/

[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 FAIL: handle_test.go:917: testStatsSuite.TestAnalyzeGlobalStatsWithOpts
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 handle_test.go:948:
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     s.checkForGlobalStatsWithOpts(c, tk, "test_gstats_opt", "p1", ca.topn, ca.buckets)
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 handle_test.go:912:
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     // since the hist-building algorithm doesn't stipulate the final bucket number to be equal to the expected number exactly,
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     // we have to check the results by a range here.
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     c.Assert(numBuckets >= buckets-delta, IsTrue)

[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 ... obtained bool = false


[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 ----------------------------------------------------------------------
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 FAIL: handle_test.go:956: testStatsSuite.TestAnalyzeGlobalStatsWithOpts2
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 handle_test.go:967:
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     s.checkForGlobalStatsWithOpts(c, tk, "test_gstats_opt2", "p1", 1, 50)
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 handle_test.go:912:
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     // since the hist-building algorithm doesn't stipulate the final bucket number to be equal to the expected number exactly,
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     // we have to check the results by a range here.
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01     c.Assert(numBuckets >= buckets-delta, IsTrue)
[2021-08-18T06:21:01.447Z] 2021/08/18 14:21:01 ... obtained bool = false

[github-actions]

Please check whether the issue should be labeled with 'affects-x.y' or 'backport-x.y.z',
and then remove 'needs-more-info' label.

[ekexium]

TestAnalyzeGlobalStatsWithOpts2 hasn't been fixed and is still unstable.

[unconsolable]

Another occurrence https://ci.pingcap.net/blue/organizations/jenkins/tidb_ghpr_check_2/detail/tidb_ghpr_check_2/35312/pipeline/62/

[2021-09-23T16:30:32.555Z] FAIL: handle_test.go:917: testSerialStatsSuite.TestAnalyzeGlobalStatsWithOpts
[2021-09-23T16:30:32.555Z] 
[2021-09-23T16:30:32.555Z] handle_test.go:948:
[2021-09-23T16:30:32.555Z]     s.checkForGlobalStatsWithOpts(c, tk, "test_gstats_opt", "p1", ca.topn, ca.buckets)
[2021-09-23T16:30:32.555Z] handle_test.go:908:
[2021-09-23T16:30:32.555Z]     c.Assert(len(tk.MustQuery(fmt.Sprintf("show stats_topn where table_name='%v' and partition_name='%v' and is_index=%v", t, p, isIdx)).Rows()), Equals, topn)
[2021-09-23T16:30:32.555Z] ... obtained int = 0
[2021-09-23T16:30:32.555Z] ... expected int = 1
[2021-09-23T16:30:32.555Z] 
[2021-09-23T16:30:32.555Z] OOPS: 114 passed, 1 skipped, 1 FAILED

[tiancaiamao]

A new occurrence in #30066

https://ci.pingcap.net/blue/organizations/jenkins/tidb_ghpr_check_2/detail/tidb_ghpr_check_2/49047/pipeline

[2021-11-30T02:29:27.292Z] FAIL: handle_test.go:921: testSerialStatsSuite.TestAnalyzeGlobalStatsWithOpts
[2021-11-30T02:29:27.292Z] 
[2021-11-30T02:29:27.292Z] handle_test.go:952:
[2021-11-30T02:29:27.292Z]     s.checkForGlobalStatsWithOpts(c, tk, "test_gstats_opt", "p1", ca.topn, ca.buckets)
[2021-11-30T02:29:27.292Z] handle_test.go:916:
[2021-11-30T02:29:27.292Z]     // since the hist-building algorithm doesn't stipulate the final bucket number to be equal to the expected number exactly,
[2021-11-30T02:29:27.292Z]     // we have to check the results by a range here.
[2021-11-30T02:29:27.292Z]     c.Assert(numBuckets, GreaterEqual, buckets-delta)
[2021-11-30T02:29:27.292Z] ... compare_one int = 0
[2021-11-30T02:29:27.292Z] ... compare_two int = 9

[qw4990]

All failed cases are caused by numBuckets or numTopN is 0, so it seems the statistics is not loaded after analyzing.

[github-actions]

Please check whether the issue should be labeled with 'affects-x.y' or 'fixes-x.y.z', and then remove 'needs-more-info' label.