In order to catch regressions it would be a good idea to create a basic github actions workflow that would trigger whenever one of our tiered storage branches changes.
What still needs to be decided upon is which tests should be run. Due to how flaky the Kafka tests are (and also how weak Github provided CI instances are) my personal preference would be explicitly pick the tests to run, specifically
All tests that have been added to test TS
Any test that has been touched by the TS feature
We can always expand on this for later but so far every bug we have found has actually been found by the specific TS unit/integration tests.
In order to catch regressions it would be a good idea to create a basic github actions workflow that would trigger whenever one of our tiered storage branches changes.
What still needs to be decided upon is which tests should be run. Due to how flaky the Kafka tests are (and also how weak Github provided CI instances are) my personal preference would be explicitly pick the tests to run, specifically
We can always expand on this for later but so far every bug we have found has actually been found by the specific TS unit/integration tests.