Closed diegodelemos closed 5 years ago
Actually this endpoint is never used. I don't see any calls to it from other REANA modules.
We did a workaround because of time constraints when reimplementing workflow stop with the independent workflow engines. We agreed to delete jobs (only Kubernetes) from RWC knowing that it would not work when we wanted to support multiple backends.
Previously this was implemented as a Celery task in REANA-Commons and it was called from RWC, which in the workflow engines was calling RJC.
We can leave this issue for moment in which we work on HTCondor and SLURM, by then we will definitely change this approach, and delete
should be generic and fully working.
Currently delete only deletes Kubernetes jobs, see here. By requiring
backend
in the incoming request we can determine which job backend class to instantiate and use the right implementation.