Add POMS campaign ID, Stage ID, Submission ID to SAM
A analysis user asked below features from project-py:
" Ideally, a metadata field indicating some sort of POMS campaign ID and submission ID should be added to files declared to SAM. Users should be able to look up that POMS campaign and submission in the POMS Web UI and find all the related information such as FHiCL files used, dataset processed etc.
Marc answered it from POMS:
We will need to put what Marc suggested in fife_launch config template.
- Target version changed from v1_2_0 to v2_0_0
You asked us to add POMS campaign and submission info into SAM metadata a while ago. So we created this issue and it is resolved now. A few weeks ago we also discussed if data generated by analysis jobs needed to be registered to SAM. I recalled the conclusion was no to most user cases. So I wonder if you were meant to register the production jobs' campaign and submission info to SAM while you asked? Could you clarify this? BTW, currently we have Project-py for analysis jobs.
#7 Updated by Pengfei Ding 3 months ago
I think the analyzers do not need their final output files registered to SAM. But if their job involves multiple stage, the output files from intermediate stage might need to be declared to SAM to form a dataset as input for the next stage jobs.
The difference between analysis jobs and production jobs is in the final stage only.
I think all files declared to SAM from a POMS campaign should have the corresponding POMS info in the metadata.
- Status changed from Accepted to Closed
We just want to make sure if the final output from a analysis job needs registering to SAM. The outputs generated during the processing jobs and as the inputs of the next stage are handled by poms. And this is internal to POMS. If one has special requests on these metadata, we can open an issue on POMS.
So as an conclusion, Project-py continues with no SAM registration for final output from analysis jobs. We will keep out cfg templates as they are now.