Hi all, I am running the pipeline and I notice from Seqera that it's been stalled at the alignment step of one sample and is trying it a second time. I then go to look at the QC the sequencing company did for me and realise that sample might have a poor quality/contaminated thus not aligning. I am just wondering if the pipeline would stop trying after a few trials and just remove the sample from the pipeline afterwards, or if I should stop the pipeline and resume it with the said sample deleted from the sample sheet. Many thanks for your help!

If a process is being rerun, there was an error that killed the first attempt. You should find out what that error was and act accordingly. E.g., if the process ran out of time or RAM and is being retried with more resources, that's probably fine. If you come to the conclusion from the QC data you have that a sample is absolute garbage and unusable, then it makes sense not to waste resources analyzing it, and it's best to remove it from the sample sheet. But if you are not sure, try to run it, and the pipeline will give you more QC information that you can use to decide if using the results is sensible.

Thanks Thomas! It was a run-time error (140). I think it runs out of the max time we can get from each node. It might be best for me to remove it for now. Many thanks!

Exit code

140

usually means that a process exceeded the allocated time. You would have to look if that was the maximum permitted on your system already, or some lower value that the pipeline used, because it is sufficient for most samples. The pipeline is typically configure to retry with more resource (doubling on first retry RMA, CPUs, and time, up to the allowed maximum governed by the

max_*

parameters). You could also explicitly modify the resource requests for a specific process in a config file. You can certainly focus on your other samples for now. The rnaseq pipeline processes samples independently from each other for the most part, so you can run the separately, if needed, and later combine the tables (if you use those).