Task #9655:  changes for rollout 3.14, missing points and changes to put setup stuff in datapoints.dpl files also incase it is datapoint dependant

Task #9607: always set status to completing befor aborted/finished, which solves the aborted trigger race condition

Annotate jobs with their project and task names to show them in summaries. Lower max job runtime to a week.

Fix nrCoresPerTask to 21, since the user or MoM not specify it correctly in all cases, and we want to use full nodes for now anyway.

Task #9607: fixed race condition. When otdb listener is triggered by onTaskScheduled, then you need to get the task status from otdb and not radb where the status still may need to be propagated to

Task #9682: Fix SLURM job status if failed, and avoid requeueing of pipeline job on node failure because SLURM messes up its administation (mixes old and new job, makes job unfindable by jobname, triggers abort-trigger job anyway)

Task #8415: Use docker-run-slurm.sh script, to contain resources and propagate kill signals in SLURM

Task #9682: Dont pull docker images from nexus anymore, since it spams the logs and crashes nexus. We distribute the images after build for now.

Task #9682: Put feedback and parsets in /data/parsets to allow the get_metadata step to rerun succesfully

Task #9682: Let SLURM jobs result in FAILED if pipeline fails, but do not run abort-trigger if pipeline could set the ABORTED status itself

Task #9682: set aborted after runPipeline.sh as well, to avoid having to wait for the follow-up job to be scheduled in case of a graceful abort

Task #9683: fixed race condition. set status to queue before we hand over the control of the pipeline to slurm