Task

The cluster has been quite clogged in this period in early 2025 and @Karl-Svard in prodbioinfo notified me about a job in balsamic cadd_annotate_somaticINDEL_research which was hardly using any threads or memory on the node on ganglia but it required 36 cores and booked up the whole node. It seems that this rule doesn't at all require this amount of resources and the cluster could be free:d up a bit if we lowered it.

I then looked at other similar rules and saw a few bcftools commands that was also run on the whole node. I don't think this should be necessary at all since the sizes of the VCFs are rarely even in the scale of 1gb.

On top of this the benchmark files specified in the rules also had the same name and should overwrite each other, meaning that we cannot track the benchmark of these rules.

Suggested approach

Lower the threads to 4 for the majority of the rules and give unique names to the rules. Then test if the cases can pass, and look into the benchmark files to see if they were at any point at risk of crashing, to help determine if 4 is a reasonable thread to posit.

Deviation

No response

Risk assessment

• Needed
• Not needed

Risk assessment link

No response

System requirements assessed

• Yes, I have reviewed the system requirements

Requirements affected

No response

Can be closed when

No response

Blockers

No response