Conversation
|
@bnaecker Do you mind taking a look at why https://github.com/oxidecomputer/dendrite/pull/997 isn't passing the deploy test? My dendrite commit requires #5593 to pass, so I'm expecting the deploy job to fail here. I'm planning to update dendrite on my PR directly. |
|
I’m not sure, but it appears the zone network setup service is absent, which is a dependency of both tfportd and dpd. Without that, I imagine making requests to dpd won’t work very well! See the log starting below. Check Run: helios / deployeng.oxide.computerMessage ID: ***@***.***>
|
|
My mail client seems to have garbled the link, here it is plain.
https://buildomat.eng.oxide.computer/wg/0/details/01J34J2AHWPZ9G5CNG3224FP96/pSdfnhxe06UxIq4Qbrpe7xA30W69Zv0SHxA3sXpY805uwbEe/01J34J2SR05BC1WEG9YR3MKEPX#S1171
|
|
@karencfv The log from the failing check updating to dendrite#997 can be found here. It indicates that Nexus didn't come up in the timeout we've set of 5 minutes. Dendrite came up fine. I would probably call this a flake. We've seen it a few times before, though we don't have the luxury of re-running it because we've already moved Dendrite past it. The next update, to fb571dc6512b24a777c5a9b2927a50501f6be297, failed here. As I mentioned yesterday, this appears to be failing because All that said, this PR is a convenience. If you are planning to update the pinned Dendrite commit that Omicron points to in your PR, I would just close this PR and do that yourself. |
Thanks! Ah, ok. So this flake has been seen before then. I think you can still run the deploy job even if the PR has moved on. I'm attempting to run the deploy job from that commit again here https://github.com/oxidecomputer/omicron/runs/27723024220.
Yes, like I mentioned in #6119 (comment) , I was expecting my commit to fail here because of the aforementioned dependencies and I had already planned to close this PR and update dendrite on my PR directly. |
|
superseded by #5593 |
Updated dendrite to commit fb571dc.