Open dmitshur opened 1 year ago
The likely explanation here is that the k8s cluster was undergoing an upgrade, resulting in brief windows of service unavailability.
Let's reuse this issue to catch future instances, so that watchflakes doesn't keep opening new issues per package/test.
Found new dashboard test flakes for:
#!watchflakes
post <- repo == "build" && `https:\/\/maintner\.golang\.org\/.+: 502 Bad Gateway`
Issue created to collect instances where internet-facing tests fail with errors: