-
Notifications
You must be signed in to change notification settings - Fork 40
LookML Validation facing error #748
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Hi @danitp97. This looks like it is as a result of an HTTP request error. Typically these are transient. Are you getting this repeatedly on this PR? |
I'll add that I'm getting this as well, running Spectacles CLI from GitHub actions. This also looks similar to a previous issue, #725 ... Seems like the same setup as above: Running something like
This generally resolves on retry for me, though this seems to be happening with increasing frequency. |
Hey! Yes, exactly I tried several times and finally worked out. Thanks!! |
Hi this is happening intermittently on our Spectacles CLI github actions. Just started popping off about 3 weeks ago and was totally fine before that. This does not resolve on retry. |
One thing I noticed is that every time (at least, every time I've looked at) we get this particular error, the GitHub action run duration is just over 5 minutes. Seems to me like there's some timeout somewhere -- and this also doesn't explain why it runs quickly (~10 seconds) without issue other times. |
Hey, I also got this error and now it is becoming more and more unstable. Maybe 1/10 could be finished and when success, it is very quick, however the rest runs until timeout. Anyone has any hacks or workarounds that I can use? |
@Hoanglinh1201 one workaround that seems to work consistently for us is, if I go to the Spectacles user in Looker that has api credentials and if I |
thank @camtr0n for the suggestion. I tried but it doesn't seem to solve our problem. I feel we cannot perform lookml validation at all now with 100/100 jobs timed out. |
Out of curiosity, how complex are everyone's LookML projects who're experiencing this? We've got a pretty large and messy project (lots of Based on the behaviour @camtr0n is describing, and our own experience, my hypothesis is that LookML validation is just taking longer than 5 minutes, and there's a rogue timeout somewhere in the client code for that long (even though the timeout is set to 7200 seconds...). LookML validation eventually finishes, though, so at some point when we retry the Spectacles run, we get a cache hit and so a retry eventually succeeds and returns results in seconds. I think this problem is worse when there are lots of CI runs being initiated, all on different PRs on different change sets. In that case, you never get a cache hit because you're always validating a different LookML configuration. We certainly find that retries are likelier to succeed when there are fewer people working on the repo simultaneously. Any ideas on how we can validate this? |
We're seeing the same issue on validations taking > 5 minutes. Here's a full stack trace from Spectacles (v2.4.8):
|
Hey everyone! Just an update that we should be pushing some update retry and httpx code shortly that should hopefully tackle some of these read errors. We don't yet have a solution to the long running LookML validation via the API, but continue to work with the team at Google/Looker on this. |
Hi,
I'm facing this error. I didn't have any issue in previous PRs. I'm not sure what does the error mean.
Could somebody help?
Thx!
The text was updated successfully, but these errors were encountered: