Allowing _cat/allocation to return more than 0 shards in test#84539
Allowing _cat/allocation to return more than 0 shards in test#84539masseyke merged 6 commits intoelastic:masterfrom
Conversation
|
Pinging @elastic/es-data-management (Team:Data Management) |
|
I'm adding |
| $body: | | ||
| /^ | ||
| ( 0 \s+ | ||
| ( \d \s+ #usually 0, unless some system index has been recreated before this runs |
There was a problem hiding this comment.
What's strange to me is that we do this above in the "all nodes" test with a \d+, but then in the "column headers" test we also assert that it's 0, so I think we'd need to change that one also if we went with this fix.
There was a problem hiding this comment.
Good point. I added that. What's going to be trickier (and maybe make us rethink these tests) are failures like #84765. That's doing a call to _cat/indices, and failing because it's getting back an unexpected second row for the index .tasks.
|
@elasticmachine update branch |
…/elasticsearch into fix/cat-allocation-test-failure
|
I think I've updated the tests for |
dakrone
left a comment
There was a problem hiding this comment.
LGTM, thanks for fixing these!
…c#84539) We delete all indices in a cluster before running ClientYamlTestSuiteIT. However some system indices can regenerate. The cat.allocation, cat.indices, and cat.shards tests were failing because they assume that no indices not created by the tests exist. They can now all handle system indices being created during the test. Closes elastic#83719 Closes elastic#84765 Closes elastic#84972 Closes elastic#82151 Closes elastic#82660
…c#84539) We delete all indices in a cluster before running ClientYamlTestSuiteIT. However some system indices can regenerate. The cat.allocation, cat.indices, and cat.shards tests were failing because they assume that no indices not created by the tests exist. They can now all handle system indices being created during the test. Closes elastic#83719 Closes elastic#84765 Closes elastic#84972 Closes elastic#82151 Closes elastic#82660
#85342) We delete all indices in a cluster before running ClientYamlTestSuiteIT. However some system indices can regenerate. The cat.allocation, cat.indices, and cat.shards tests were failing because they assume that no indices not created by the tests exist. They can now all handle system indices being created during the test. Closes #83719 Closes #84765 Closes #84972 Closes #82151 Closes #82660
We delete all indices in a cluster before running ClientYamlTestSuiteIT. However some system indices can regenerate.
The cat.allocation, cat.indices, and cat.shards tests were failing because they assume that no indices not created by the tests exist. They can now all handle system indices being created during the test.
Closes #83719
Closes #84765
Closes #84972
Closes #82151
Closes #82660