graphite having tons of issues keeping GHA checks ...
# report-bugs
j
graphite having tons of issues keeping GHA checks up to date on the UI. can someone please help us investigate, it's really brutal. We're close to churning graphite
p
(If applicable) To help us repro, please send the following if you haven't already: - PR links, screenshots, and/or debug output if applicable - Repro steps - Does this happen consistently? Just to you or others on your team also? - Is this a regression? Is the problem new? - If you're having an issue with the CLI, please send us debug logs using
gt feedback --with-debug-context <any message>
j
image.png
<https://app.graphite.com/github/pr/Rogo-Technologies/ask-rogo/32655>
this is what it shows on github
This has been happening on dozens of PRs the last couple of days
p
hey Jan, could you please have a look for that
Typescript CI Status
 in the PR in GitHub just to confirm it is successful there. Running the logs for your PR on our end
j
hey @Gareth Brickman yes it's successful on github, see the screenshot from github
p
wanted to see that check specifically. Looks like it's actually a suite of checks, I think that could be important for the underlying issue of why it's mis-reporting in Graphite
j
yep it is
sorry I'm trying to be more helpful. To be honest we've grown a bit frustrated with graphite over the last couple weeks, we've just had so many issues where GHA data got out of sync with what's on GitHub.
let me find the link
okay so I ended up closing that PR and then reopening it, which caused all CI to get rerun and Graphite's cache to get cleared / data to get synced again. Unfortunately, I don't think I'll be able to find that exact run anymore
This PR had the same issue with the Cursor bugbot being reported as "pending" while it was showing as completed successfully on github https://app.graphite.com/github/pr/Rogo-Technologies/rogo-infra-common/3377/Added-MT-SG-to-the-shared-reducto-PSC
p
okay, now I know this particular case is a check suite I'll make my own one and try reproduce that! I suspect while the check suite is in fight a push to the stack invalidates its initial run and starts a new run but for whatever reason Graphite becomes unaware there is a newer run that should be monitored. Apologies for the frustration here, we also encounter it in our PRs with GHA and the flakeyness and inherent issues with how they report state is something we're grappling with
j
thank you!!
d
We've had a similar issue for the last couple weeks, where Cursor Bugbot specifically is constantly marked as "Reviewing your changes…" even after it finishes running. On Github it properly shows up as having completed. Really frustrating since it's a required check in our repo and it not being marked as completed on the Graphite side makes us unable to merge on graphite and take advantage of auto-restack etc
j
Hey @Gareth Brickman so I found another instance of this on this PR right now: https://app.graphite.com/github/pr/Rogo-Technologies/ask-rogo/33021
👀 1
g
gotcha, I'm going to see about getting Cursor Bugbot set up in my sandbox GitHub org to try reproduce this behavior