I have a docker build job that says it took 46 seconds, but that’s because it’s the second attempt and it used the cache from the first step which timed out after 20 minutes. I think the canvas UI used to show to sum of all attempts.
I’m even more confident that failed attempts showed up as red bars on the waterfall UI just a few days ago and that’s no longer the case.
I thought the canvas UI used to show the sum of durations, even though it didn’t explicitly show the attempts.
Looking closer at my waterfall, though, it looks like my expectation is maybe true in general, but not true for timeouts?
The attached screenshots are all from the same build. “ :aws build” timed out on the first attempt and succeeded on the second, yet only the second attempt appears on the waterfall. farther down, though, “ test-integration” has a red attempt followed by a green attempt because its tests intentionally fail on the first run.
Retry visibility should be the same regardless of the timeout. I tried reproducing your scenario, but have not been successful so far. Could you share your build link to support@buildkite.com for us to investigate further?