fix resubmission tests #78
Conversation
on the 2nd submission the env variables get lost. For some reason bash evaluates this as successful even if the if statement crashes. so actually there are only 2 submissions (1 resubmission), i.e. multiple resubmissions still do not work. Hence I suggest to to a "real OOM" resubmission and treat the case of a lost variable as error.
Pull Request Test Coverage Report for Build 3923746443Warning: This coverage report may be inaccurate.This pull request's base commit is no longer the HEAD commit of its target branch. This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.
Details
💛 - Coveralls |
|
@bernt-matthias Thanks for making this PR. @cat-bro and I were discussing this issue, and ultimately ,because resubmission doesn't work with dynamic destinations, it looks like we would need some kind of fix in Galaxy for things to work at an acceptable level. One possibility would be to somehow remove the cache in resubmits only, and recompute the destination afresh. Is that your understanding as well? |
There was a problem hiding this comment.
@bernt-matthias Thanks for fixing this test. I'm hoping something can be fixed on the Galaxy end at some point, but till then I'll rebase the existing tests for activating resubmit against this.
on the 2nd submission the env variables get lost: see galaxyproject/galaxy#9747 .. galaxyproject/galaxy#12852
For some reason bash evaluates this as successful even if the if statement crashes. so actually there are only 2 submissions (1 resubmission), i.e. multiple resubmissions still do not work.
Even if the tests are not executed at the moment I would suggest to fix them by doing a "real OOM" resubmission and treat the case of a lost variable as error.