A community effort to collect a curated set of analysis pipelines built using Nextflow.

nf-core

Hi, I posted this on <#CE8SSJV3N|>, was redirected here, this seems to be the right forum. New to all of this. I'm running the nf-core rnaseq pipeline on Uppmax/Rackham for the first time, just with 2 samples to begin with to test it, and it hasn't run out of time yet, but it seems that multiqc was killed. I don't understand why.

.nextflow.log

140 would typically be `SIGUSR2` which is most commonly received to give jobs a chance to gracefully shut down when they are due to be killed because of their time being up.

What version of nextflow are you running? It should typically restart with more resources (longer time). Since it's stopped, you may consider tuning the process to give it more time (see <https://nf-co.re/docs/usage/getting_started/configuration#tuning-workflow-resources>)

Ok, so a resource issue, that makes sense. Running the latest version, 24.10.5.5935.

Is it this issue? <https://community.seqera.io/t/multiqc-timing-out/2074>
If so you might have to wait for the fix, or change the multiqc configuration not to export plots. The pipeline should be able to pass a second config to multiqc.

If that was the issue, then it should work now, right? Looks like a fix has been released. I'll give it another go.

I don't believe it's in the current release version yet. Only the main branch of multiqc

i.e. you would need to install MultiQC in your PATH and supply
```process {
    withName: MULTIQC {
        container = null
    }
}```
and then run the workflow with the custom config

Yeah, you're right of course. I'll try that, thanks for your help!

Looking more into that, it seems to be occuring randomly? In that case, rerunning with `-resume` would be the suggested path forward for now (possibly checking up on the multiqc jobs, and killing them if they just hang).

Yes, good point, I don't need to start again from scratch.

It got stuck again at MultiQC, but a new job with slightly different parameters (none of which should affect the MultiQC step) ran through, so it does indeed seem to be random.