I think we need to be practical: We either need to take the time to improve our test and build systems properly, or we reduce the flakyness test by test - the first is much better, but we currently we are doing neither.
I <b>100%</b> think that we need to take the time to go deep into this. But while we don't, our builds fail more than 50% of the time due to 1 or 2 flaky tests failing. This often brings hours (if not days, sometimes) of delay in deployments. FWIW, I ran this test with a 0.2 second timeout locally with `--repeat 20` and it always ran successfully (it fails if I set the timeout to 0.1). But when the buildbot runs it, sometimes it somehow takes longer than 30 seconds. Also FWIW, I don't expect this to add 30 seconds to the test run - it only fails from time to time, so it probably takes only slightly longer than the 30 seconds. So although this is not the best solution, it's at least something. Happy to discuss this at stand-up tomorrow! -- https://code.launchpad.net/~ines-almeida/launchpad/+git/launchpad/+merge/458930 Your team Launchpad code reviewers is requested to review the proposed merge of ~ines-almeida/launchpad:increase-unittest-fetch-timeout into launchpad:master. _______________________________________________ Mailing list: https://launchpad.net/~launchpad-reviewers Post to : launchpad-reviewers@lists.launchpad.net Unsubscribe : https://launchpad.net/~launchpad-reviewers More help : https://help.launchpad.net/ListHelp
