Commit Graph

266 Commits

Author SHA1 Message Date
Antonio Andelic
c60d9db687
Merge branch 'master' into ignore-attach-thread-keeper-errors 2022-09-26 08:38:48 +02:00
Antonio Andelic
5ff1bcd553
Merge branch 'master' into collect-correctly-logs-in-stress-test 2022-09-26 08:38:38 +02:00
alesapin
06e0f554d8 Fix fetch to local disk 2022-09-23 16:46:53 +02:00
Antonio Andelic
1d93c56d1a Collect logs using clickhouse-local 2022-09-23 10:54:16 +00:00
Antonio Andelic
a17a3e1de1 Ignore Keeper hardware errors 2022-09-23 08:23:57 +00:00
kssenii
46f74aaba9 Update stress/run.sh 2022-09-12 20:10:35 +02:00
Alexander Tokmakov
e77b9e4d0c
Merge pull request #40775 from azat/ci/core-dumps-rework
Rework core collecting on CI (eliminate gcore usage)
2022-09-09 20:20:10 +03:00
Alexey Milovidov
7776512b04
Merge pull request #41002 from azat/ci/fix-oom-check
ci/stress: clear dmesg before run to fix "OOM in dmesg" check
2022-09-06 06:41:36 +03:00
Alexander Tokmakov
b264be3c63
Merge branch 'master' into zookeeper_client_fault_injection 2022-09-05 22:13:09 +03:00
Azat Khuzhin
2724b67537 ci/stress: clear dmesg before run to fix "OOM in dmesg" check
CI: https://s3.amazonaws.com/clickhouse-test-reports/40772/afa137ae2b6108e72c2d6e43556a04548afa2ea9/stress_test__ubsan_.html
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-09-05 15:51:36 +02:00
Azat Khuzhin
25e3bebd9d Rework core collecting on CI (eliminate gcore usage)
gcore is a gdb command, that internally uses gdb to dump the core.

However with proper configuration of limits (core_dump.size_limit) it
should not be required, althought some issues is possible:
- non standard kernel.core_pattern
- sanitizers

So yes, gcore is more "universal" (you don't need to configure any
`kernel_pattern`), but it is ad-hoc, and it has drawbacks -
**it does not work when gdb fails**. For example gdb may fail with
`Dwarf Error: DW_FORM_strx1 found in non-DWO CU` in case of DWARF-5 [1].

  [1]: https://github.com/ClickHouse/ClickHouse/pull/40772#issuecomment-1236331323.

Let's try to switch to more native way.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-09-04 22:07:16 +02:00
Alexander Tokmakov
8bdb589c2b Merge branch 'master' into zookeeper_client_fault_injection 2022-08-29 13:34:57 +02:00
alesapin
133ca01447 Merge branch 'master' into stress_s3 2022-08-29 11:25:28 +02:00
Azat Khuzhin
ebc61a36e0 tests/stress: improve OOM detection (add separate check by dmesg)
Right now if you will look at the OOM errors:
- OOM killer (or signal 9) in clickhouse-server.log
- Backward compatibility check: OOM messages in clickhouse-server.log

Most of them are not real, but just clickhouse server got KILLed by
clickhouse stop, #40678 may imporove the situation, but to definitely
sure that there was OOM let's look at dmesg.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-08-27 12:46:58 +02:00
Azat Khuzhin
3b519c5d44 tests/stress: capture stacktrace of server hungs if pid was removed already
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-08-26 22:06:32 +03:00
alesapin
704c4b2c5b Stop thread fuzzer on shutdown 2022-08-26 11:54:54 +02:00
alesapin
3ff6489fae Merge branch 'master' into stress_s3 2022-08-25 13:14:58 +02:00
alesapin
ad692f732a Merge branch 'master' into stress_s3 2022-08-25 13:13:30 +02:00
alesapin
35f9815b8e Fix backward comp check 2022-08-24 14:43:02 +02:00
Azat Khuzhin
50bddc43dc tests/stress: ignore NETLINK_ERROR from checkPermissionsImpl
Since now with --privileged it has CAP_SYS_ADMIN and tries to
communicate via netlink.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-08-19 14:07:06 +02:00
Alexander Tokmakov
3d253ec51b
Merge branch 'master' into zookeeper_client_fault_injection 2022-08-18 21:23:50 +03:00
alesapin
922818bbd9 Merge branch 'stress_s3' of github.com:ClickHouse/ClickHouse into stress_s3 2022-08-18 14:46:50 +02:00
alesapin
932ea146f5 Merge branch 'master' into stress_s3 2022-08-18 13:14:47 +02:00
alesapin
86b1e33eed Disable cache on writes 2022-08-17 19:00:53 +02:00
alesapin
600d22851f Grep dangerous S3 errors 2022-08-17 12:43:11 +02:00
alesapin
0433b801d2 Configure properly 2022-08-17 12:27:15 +02:00
Alexander Tokmakov
ae000e9125
Merge branch 'master' into zookeeper_client_fault_injection 2022-08-17 12:48:54 +03:00
alesapin
1ec6627a70 Fix tables creation 2022-08-16 18:28:17 +02:00
kssenii
eb26b219b9 Merge master 2022-08-16 00:56:27 +02:00
Alexander Tokmakov
589c3408d2
Merge pull request #40234 from ClickHouse/better_message_on_restore_covered
Better error message when restoring covered parts
2022-08-15 22:01:48 +03:00
alesapin
243bd492fa Trying to fix it 2022-08-15 20:55:11 +02:00
alesapin
96722a13bb Merge branch 'master' into stress_s3 2022-08-15 20:20:31 +02:00
Alexander Tokmakov
edaff70010 better error message when restoring covered parts 2022-08-15 13:53:14 +02:00
Alexander Tokmakov
467ef7bbc2
Update run.sh 2022-08-12 14:30:18 +03:00
Alexander Tokmakov
b9d18182f2 fix 2022-08-11 15:27:26 +02:00
kssenii
5c3227ba56 Merge master 2022-08-10 12:00:34 +02:00
kssenii
0dda03c94b Fix checks 2022-08-10 00:06:58 +02:00
Azat Khuzhin
3772415588 tests/stress: add dmesg output (to see OOM details)
max_server_memory_usage already set to 75%, so OOM should not happens,
the reason is that because RSS does not match with memory tracker
statistics:

    2022.08.05 12:36:57.869896 [ 82524 ] {} <Trace> AsynchronousMetrics: MemoryTracking: was 64.69 GiB, peak 65.26 GiB, will set to 62.80 GiB (RSS), difference: -1.89 GiB
    ...
    2022.08.05 12:37:00.213440 [ 82334 ] {} <Error> void DB::MergeTreeBackgroundExecutor<DB::MergeMutateRuntimeQueue>::routine(DB::TaskRuntimeDataPtr) [Queue = DB::MergeMutateRuntimeQueue]: Code: 241. DB::Exception: Memory limit (total) exceeded: would use 64.68 GiB (attempt to allocate chunk of 1298794 bytes), maximum: 51.44 GiB. OvercommitTracker decision: Memory overcommit isn't used. Waiting time or orvercommit denominator are set to zero.. (MEMORY_LIMIT_EXCEEDED), Stack trace (when copying this message, always include the lines below):

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-08-06 12:34:38 +03:00
kssenii
7a9b0bc47f Merge master 2022-08-05 01:48:52 +02:00
kssenii
d462782d1a Fix checks 2022-08-02 14:27:45 +02:00
Alexander Tokmakov
e5c47cb26f
Update run.sh 2022-08-02 12:10:53 +03:00
Alexander Tokmakov
ecf7ce1f74 Merge branch 'master' into zookeeper_client_fault_injection 2022-08-01 20:49:01 +02:00
Alexander Tokmakov
3cc20f05ba
Update run.sh 2022-08-01 20:47:14 +03:00
kssenii
e5f4a619ed Merge master 2022-07-31 20:24:40 +03:00
Mikhail f. Shiryaev
c6b251d0cb
Revert "Merge pull request #38106 from ClickHouse/tsan_less_flaky_tests"
This reverts commit 0ae0cc54aa, reversing
changes made to 1d7cf28cab.
2022-07-22 12:14:23 +02:00
Alexander Tokmakov
840ace663b
Update run.sh 2022-07-20 13:19:53 +03:00
kssenii
8ad26b3111 Merge master 2022-07-16 22:50:03 +02:00
Alexander Tokmakov
b770f9903d
Update run.sh 2022-07-15 20:03:00 +03:00
Alexander Tokmakov
5c24cd979b
Merge pull request #39256 from ClickHouse/bc_check_fix_stacks_in_gdb
Fix stacktraces in gdb in BC check
2022-07-15 17:54:18 +03:00
Alexander Tokmakov
77487e996c fix stacktraces in gdb in bc check 2022-07-15 12:23:28 +02:00
kssenii
15a2d60f23 Fix checks 2022-07-14 22:04:39 +02:00
Alexander Tokmakov
f8e23c1deb
Update run.sh 2022-07-14 17:29:08 +03:00
Alexander Tokmakov
991493e6fc
Update run.sh 2022-07-13 15:22:36 +03:00
Alexander Tokmakov
948f15eff3
Update run.sh 2022-07-12 14:01:53 +03:00
Alexander Tokmakov
2b886ddc2e
Update run.sh 2022-07-08 16:47:54 +03:00
Alexander Tokmakov
57b8a6a542
Update run.sh 2022-07-08 13:53:35 +03:00
Kruglov Pavel
8ffd6cd22c
Merge pull request #38717 from Avogar/improve-bc-check
Try to improve backward compatibility check
2022-07-07 18:52:34 +02:00
Kruglov Pavel
a3a94a1c09
Avoid false positive red sanitizer asserts check in stress test 2022-07-06 14:32:17 +02:00
Kruglov Pavel
45ac74617b
Fix 2022-07-04 20:38:21 +02:00
Kruglov Pavel
65dd81586b
Update run.sh 2022-07-04 17:30:22 +02:00
avogar
17da24b584 Try to improve backward compatibility check 2022-07-04 13:02:22 +00:00
Azat Khuzhin
8e9c1eaa73 tests/stress: redirect gdb output on stop hang to gdb.log
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-07-03 14:46:17 +03:00
Alexander Tokmakov
b238ed4093 fix stress tests 2022-07-01 17:19:00 +02:00
Alexander Tokmakov
3ff26939fc
Update run.sh 2022-06-29 15:46:40 +03:00
Alexander Tokmakov
428628238f Merge branch 'master' into deprecate_ordinary_database 2022-06-28 15:40:33 +02:00
Kruglov Pavel
25100ad5e5
Merge pull request #38520 from ClickHouse/Avogar-patch-1
Fix application errors grep in stress test
2022-06-28 14:26:22 +02:00
Alexander Tokmakov
411512ed6f
Update run.sh 2022-06-28 14:28:06 +03:00
Kruglov Pavel
f7419e0703
Fix application errors grep in stress test 2022-06-28 13:12:21 +02:00
Alexander Tokmakov
dbf2763788 automatically convert system database to Atomic 2022-06-23 21:38:43 +02:00
Azat Khuzhin
dce71bbe06 tests/stress: fix TSan detection
+ test -f package_folder/clickhouse-client_22.7.1.1+tsan_all.deb package_folder/clickhouse-common-static-dbg_22.7.1.1+tsan_amd64.deb package_folder/clickhouse-common-static_22.7.1.1+tsan_amd64.deb package_folder/clickhouse-keeper_22.7.1.1+tsan_amd64.deb package_folder/clickhouse-server_22.7.1.1+tsan_all.deb
    /run.sh: line 13: test: too many arguments

Follow-up for: #38207 (cc @tavplubix)
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-06-20 08:10:03 +03:00
Azat Khuzhin
d154288fc6 tests/stress: fix TSan detection (enables thread fuzzer for non-TSan builds)
Follow-up for: #38106 (cc @tavplubix @alesapin)
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-06-19 19:27:25 +03:00
alesapin
0ae0cc54aa
Merge pull request #38106 from ClickHouse/tsan_less_flaky_tests
Some fixes for tests with tsan
2022-06-18 14:22:43 +02:00
Alexander Tokmakov
3e8aa2dc47
Update run.sh 2022-06-16 13:24:21 +03:00
Alexander Tokmakov
9cac78b498 make tests with tsan less flaky 2022-06-15 19:54:46 +02:00
Mikhail f. Shiryaev
4d45839f17
Generalize setup_minio.sh
- Cleanup apt garbage in the container
- Download the same binary in docker and setup_minio.sh
- Use binary and not rpm for minio
- Sort packages in Dockerfile's
2022-06-10 15:04:45 +02:00
Kruglov Pavel
83418ebba2
Merge pull request #37712 from vdimir/duplicate_download_previous_release
Get rid of duplicate download_previous_release
2022-06-07 12:51:34 +02:00
Kruglov Pavel
196863ade5
Ignore PART_IS_TEMPORARILY_LOCKED error in BC check 2022-06-02 13:33:27 +02:00
Vladimir C
2622791b9e
upd 2022-06-01 16:17:17 +02:00
mergify[bot]
3866c27850
Merge branch 'master' into stress_s3 2022-06-01 12:02:33 +00:00
vdimir
4d336d0228
Get rid of duplicate download_previous_release 2022-05-31 17:50:29 +00:00
Alexander Tokmakov
c23bc01196 Merge branch 'master' into zookeeper_client_fault_injection 2022-05-11 17:20:42 +02:00
alesapin
af99183950 Disable thread fuzzer after server restart 2022-05-10 15:16:21 +02:00
mergify[bot]
5c742cf2d5
Merge branch 'master' into stress_s3 2022-05-10 12:26:26 +00:00
Alexander Tokmakov
9346c9363a
Merge pull request #37052 from ClickHouse/fix_server_termination_stress_tests
Print stacks if we cannot terminate server in stress tests
2022-05-10 13:12:44 +03:00
Alexander Tokmakov
17c954bea7 fix shellcheck 2022-05-10 11:48:55 +02:00
Alexander Tokmakov
92d423c478 print stacks if wa cannot terminate server in stress tests 2022-05-09 19:43:51 +02:00
Alexander Tokmakov
b647f8f9bf
Update run.sh 2022-05-09 17:48:10 +03:00
Alexey Milovidov
419687ac50
Update run.sh 2022-05-09 10:17:49 +02:00
Mikhail f. Shiryaev
b14dca2fac
Improve stress tests report a little bit 2022-05-09 10:17:40 +02:00
alesapin
85dda249b8 Don't download stateful table 2022-05-02 00:11:05 +02:00
alesapin
51345fa25b Trying stress on top of S3 2022-05-01 19:07:02 +02:00
Kruglov Pavel
4e2b1cd981
Update run.sh 2022-04-20 16:48:54 +02:00
Kruglov Pavel
cd77f927b6
Fix putting logs into test_output 2022-04-20 15:06:09 +02:00
Kruglov Pavel
6bcdf21801
Fix stress test 2022-04-20 13:53:16 +02:00
Kruglov Pavel
eaad99bd48
Try to turn off always green backward compatibility checks 2022-04-08 13:55:10 +02:00
Kruglov Pavel
9fa6e6e7df
Make backward compatibility check always green 2022-04-07 17:42:33 +02:00
Azat Khuzhin
2fb6d6acb4 stress: use more meaningful names for clickhouse-server.log
But note, that stderr.log/stdout.log/clickhouse-server.err.log still not separated.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-04-06 14:09:53 +03:00
Azat Khuzhin
058cde820f test/stress: improve error checks for backward compatiblity check
Before the log was analyzed only after the server had been restarted
after stressing, and hence it cannot find anything in case of error,
like here [1].

  [1]: https://s3.amazonaws.com/clickhouse-test-reports/34355/acd48ebe2d51d20d0c2a5bc545d46d27f9bf7904/stress_test__thread__actions_.html

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2022-04-06 14:09:52 +03:00
Alexander Tokmakov
307a2ca579 Merge branch 'master' into zookeeper_client_fault_injection 2022-03-31 15:32:41 +02:00
Kruglov Pavel
18ab49e788
Check all logs for crashes, logical errors, etc in backward compatibility check 2022-03-25 11:59:50 +01:00