Open def- opened 4 months ago
Haven't seen this re-occur in a few months - am inclined to close this. Thoughts?
Sure, let's close and reopen if it happens again.
We have a similar issue in CI being flaky again: https://buildkite.com/materialize/test/builds/84628#019052ac-d24f-427e-914e-5909d8cd5680
>> ALTER TABLE alter_decimal MODIFY COLUMN f1 DECIMAL(6,1);
>> INSERT INTO alter_decimal VALUES (12345.6);
> SELECT * FROM alter_decimal;
query error didn't match; sleeping to see if dataflow produces error shortly 50ms 75ms 113ms 169ms 253ms 380ms 570ms 854ms 1s 2s 3s 4s 6s 10s 15s 16s^^^ +++
alter-table-after-source.td:297:1: error: query succeeded, but expected error containing "altered"
|
18 | > CREATE SECRET mysq ... [rest of line truncated for security]
22 | PASSWORD SECRET ... [rest of line truncated for security]
25 | $ mysql-connect name ... [rest of line truncated for security]
296 |
297 | ! SELECT * FROM alter_decimal;
| ^
--- alter-table-irrelevant.td
And another one, reported by Frank: https://materializeinc.slack.com/archives/C01LKF361MZ/p1720812756750439
Still unclear why this is flaky but it is worth noting that this test file alter-table-after-source.td
has these overrides which are not in many of the other tests
$ set-sql-timeout duration=1s
$ set-max-tries max-tries=20
It's possible that we are just not giving enough time for these errors to be detected and propagated
Let's see if that helps! https://github.com/MaterializeInc/materialize/pull/28247
Tried that, didn't help, still seeing the same error.
This is still occurring, I'll disable the test for now: https://buildkite.com/materialize/test/builds/86989#0190fd81-49e2-4588-b657-6e9f69d80677
What version of Materialize are you using?
937b9e3b0ca1
What is the issue?
Seen in https://buildkite.com/materialize/test/builds/80985#018ef6a4-f623-42bb-a362-49dc4fdbc421 with MySQL CDC
ci-regexp: alter-table-after-source.td: