Conversation
|
Thanks @j-s-135, the changes look good, but oddly the tests are failing with an error message that doesn't seem to be related, and one that @jeremy-rcsb actually encountered in his k8s workflow: I'm wondering if this PR created a new image on Harbor, which is what the k8s workflow started using? This error is very strange since my PR just two days ago didn't encounter any issues: #30 One thing though— @j-s-135—can you increment the version number in HISTORY and FYI @brindakv |
|
Ugh, I just confirmed (in an essentially empty PR) that this appears to be a real issue now, for some reason. Normally, this is what we see (from our Azure pipelines after merging code on Wednesday): But now, making no changes to the code—although, after merging @brindakv, do you have any idea what this might be related to? I'm very confused on how production worked yesterday given that my new blank PR is failing without any actual code changes. |
|
@j-s-135 @jeremy-rcsb @brindakv Alright, so after banging my head for an hour, I think I figured out the issue. I believe the PDBxLoader was trying to access the IHM dictionary based on the configuration file in mock-data, to which the IHM dictionary setting hadn't yet been added. I just added it here, and re-ran my PR workflow, and I think it is working now. @j-s-135, so, for your PR to pass, you will need to update mock data. I'll do this for you now since I just updated mock-data myself, but for the record here would be the steps (while within the # to update mock-data in the `py-rcsb_workflow` repo:
git submodule update --recursive --remote
# to check if the update was already added to your next commit
git status
# if it wasn't added yet, please add it:
git add -u
# Then commit and push the updated mock-data directory
git commit
git push |
|
@j-s-135 So the update of mock-data fixed the original issue this PR was facing, but there seems to be a new issue, possibly due to the handling of how dictionaries are read in for your BCIF workflow. It seems like the workflow is reading in the dictionary file many times in a row. Can you please review your code to see if this is the case and adjust it so that it only reads in the dictionary once? Also, before I forget, can you please update the the version in the HISTORY.txt and |
There was a problem hiding this comment.
Thanks @j-s-135, please see my last comment for change requests (#31 (comment)). Also, please remember to pull my commits before pushing your new changes.
|
I think it's fixed. The other errors seem to relate to fixturePdbxLoader. |
piehld
left a comment
There was a problem hiding this comment.
Thanks @j-s-135. Your change looks good, but I think you may have accidentally undone the updates I made to mock-data which fixed the fixturePdbxLoader issues.
Can you please run the following on your machine:
# navigate to your repo directory
cd py-rcsb_workflow
# update mock-data:
git submodule update --recursive --remote
# if it wasn't added yet, please add it:
git add rcsb/mock-data
# Then commit and push the updated mock-data directory
git commit
git push
syncing with dictionary link from exdb