Closed rvantonder closed 3 years ago
Issue cropped up here: https://github.com/sourcegraph/sourcegraph/pull/10572/checks?check_run_id=664502349
Run src lsif upload -github-token=***
Repository: github.com/sourcegraph/sourcegraph
Commit: 83a5e262b21cfd7acdcc4ec82678035dc6484f0f
File: ./dump.lsif
Root: cmd/precise-code-intel
error: 500 Internal Server Error
no precise-code-intel-api-server endpoints could be found (this may indicate more precise-code-intel-api-server replicas are needed, contact support@sourcegraph.com for assistance)
##[error]Process completed with exit code 1.
Separately, this issue cropped up with users forking a repo and Sourcegraph.com returning a 500 code because it was busy cloning the fork.
Separately, this issue cropped up with users forking a repo and Sourcegraph.com returning a 500 code because it was busy cloning the fork.
@rvantonder implementing #9700 this week which will fix that specific case, also filed #10576: re changing default behaviour of indexing on forks (for forks of public repos, don't do it by default)
https://github.com/sourcegraph/sourcegraph/pull/10914/checks?check_run_id=700718622
Repository: github.com/sourcegraph/sourcegraph Commit: 2ae4cf6e37b152e7c9c6b0f652f79a6c2c0da2d1 File: ./dump.lsif Root: browser error: 401 Unauthorized
2 errors occurred:
- given repository sourcegraph/sourcegraph not listed in installed repositories
- you do not have write permission to the repository
Related to GH being down maybe?
big sad. putting up https://github.com/sourcegraph/lsif-upload-action/pull/14 as option for external projects (don't think we should enable that on our repos, good to know about what affects uploads)
Another one, not sure why
Repository: github.com/sourcegraph/sourcegraph
Commit: 7552f28ed35a5bc2c9839e78e7f2d51dc3ce3916
File: ./dump.lsif
Root:
Post "https://sourcegraph.com/.api/lsif/upload?commit=7552f28ed35a5bc2c9839e78e7f2d51dc3ce3916&github_token=***&repository=github.com%2Fsourcegraph%2Fsourcegraph": read tcp 172.18.0.2:60140->104.26.9.187:443: read: connection reset by peer
##[error]Process completed with exit code 1.
Another
Run src lsif upload -github-token=***
Repository: github.com/sourcegraph/sourcegraph
Commit: 0ab59c2994d191be02e251ee43bc260aa9c4af8a
File: ./dump.lsif
Root: web
error: 500 Internal Server Error
git command [git rev-parse 0ab59c2994d191be02e251ee43bc260aa9c4af8a^0] failed (stderr: ""): Post "http://gitserver-4.gitserver:3178/exec": dial tcp: lookup gitserver-4.gitserver on 10.55.240.10:53: no such host
Upload LSIF data29s
##[error]Process completed with exit code 1.
Run src lsif upload -github-token=***
Repository: github.com/sourcegraph/sourcegraph
Commit: 4a68001a2d73d3aaea1ccd0c9501ca75a40dfb93
File: ./dump.lsif
Root:
error: 502 Bad Gateway
<html>
<head><title>502 Bad Gateway</title></head>
<body>
<center><h1>502 Bad Gateway</h1></center>
<hr><center>nginx/1.17.8</center>
</body>
</html>
job Probably when pods bounce
Closing due to inactivity.
When an LSIF job fails for whatever reason, the PR build at a glance fails as a whole (not all checks passed, red X). The overhead of checking whether the failure is due to a broken build (merge blocking) versus LSIF (not merge blocking) can be a pain point. Resolving this in part comes down to not having full control over the nature of GH's check logic or visual indicators.
@gbrik says (paraphrased):
Let's use this issue to track cases of LSIF failures to inform whether this is a recurrent thing worth thinking about.