PM job state management #3702
PM job state management #3702
Conversation
| @@ -103,12 +103,13 @@ def destroy | |||
|
|
|||
| # GET /projects/:project_id/jobs/:cluster/:jobid | |||
| def job_details | |||
| project = Project.find(params[:project_id]) | |||
There was a problem hiding this comment.
You should avoid using params directly and instead us maybe job_details_params so the list of parameters we're accessing are known to be valid and permitted.
| def job_from_id(job_id, cluster) | ||
| launchers = Launcher.all(directory) | ||
| launchers.each do |launcher| | ||
| job = launcher.job_from_id(job_id, cluster) | ||
| return job unless job.nil? | ||
| end | ||
| end |
There was a problem hiding this comment.
There's something wrong here that I think I started. jobs are associated with the launchers and that's why you have to do all this weirdness here. But I think a better/simpler paradigm really is that the project itself keeps the job log and the entries have a relationship to the launcher (if we even need such a relationship).
Which is really just to say that I think we should rethink & refactor this whole paradigm before it gets worse. I think we can move forward with this as it is, but will likely have to change it almost immediately.
* better pm job state management * move job search to function * use job_details_params
Fixes #3685
Updates the job log whenever a job is queried from the adapter, and does not query when the job is logged as being completed.