We have a workflow that contains 2 jobs that run in parallel. These jobs run end-to-end tests for Android and iOS platforms. A third job is intended to take files that are created and persisted via persist_to_workspace and act on them (after using attach_workspace). However, this job must run even if one of the testing jobs fails. In order to accomplish this, I’ve successfully set up the workaround to allow for running jobs sequentially regardless of the outcome of the required jobs (title of the workaround is [Workaround] Run jobs sequentially regardless of the outcome of the required jobs as I cannot post links). Note that using the workaround means removing the explicit requires dependency between the final job and the upstream ones. I believe the persist_to_workspace and attach_workspace require this relationship; at least that’s what the article (titled Using Workspaces to Share Data between Jobs would have me believe.
During the testing jobs, the following is logged when persisting:
Creating workspace archive...
Uploading workspace archive...
Total size uploaded: 251 B
Workspace archive uploaded successfully.
But during the completion job, the following is printed:
Downloading workspace layers
Total size downloaded: 0 B
Applying workspace layers
591b840f-6436-4f29-828a-0301371407f1 - persisted no files
This leads to a FileNotFoundException when trying to access the files.
Any chance I’m just missing something? Can I provide any more information?
Thanks!