StanfordAHA/lassen

Buildkite CI pipeline broken; we appear to be approving PR's even though CI is failing...?

Opened this issue · 1 comments

Our buildkite CI test https://buildkite.com/stanford-aha/lassen has been failing steadily since November 2023.
As a result, it looks as though pull requests are being merged even though this test fails.
See e.g. this recent merge: https://buildkite.com/stanford-aha/lassen

"1 of 2 checks passed" => "View details" => "buildkite/lassen Build #697 failed"

Here is an example of the failing test: https://buildkite.com/stanford-aha/lassen/builds/697

===== 950 failed, 128 passed, 1 skipped, 72 warnings in 403.26s (0:06:43) ======

We should either fix the failing pytests or else remove them from the CI test suite...
who is the primary maintainer(s) for lassen these days...?

@jack-melchert @bobcheng15 I included you guys as "assignees" since you were the most recent people to touch the code, which may be unfair...feel free to unassign yourself and/or pass the assignation along to someone else...