I have succeeded in installing nextflow and quantms - not that hard and I feel confident that the problem is not there. I have been testing the pipeline with the suggested test
nextflow run bigbio/quantms -profile test_lfq,apptainer

The pipe line will do something and eventually fail. This is an example of the error:
Error - cannot open parameter file "/tmp/20250910_181120_compute01_41_1/param.txt".
Process '/opt/OpenMS/thirdparty/Comet/comet.exe' did not finish successfully (exit code: 1). Please check the log.
There is nothing interesting in the log..
I observed that by repeating the pipeline with -resume I would progress a bit every time and eventually finish.
I have been thinking the problem is grounded in IO race conditions. I did fix this particular error (or I have not seen it repeated) by adding a 'sync' to the modules/local/openms/comet/main.nf

A new error has since popped up:
Could not remove directory 20250911_104006_compute01_42_1!
Error: File not found (the file '/tmp/20250911_104006_compute01_42_1/result.pep.xml' could not be found) The observation is the same: Repeating with -resume the pipeline will eventually finish successfully and the problem/situation can be explained by IO race condition.

I have also remounted my disk to sync option instead of async, which I believed would fix a race condition, but the difference in pipeline behavior was difficult to spot.

I am using a modern Dell server with ubuntu server 24.04.3 LTS. It is part of a cluster.
I realize that nextflow and quantms is mature software, and should not suffer from these kind of problems, so I am reaching out in the hope that this is a known issue, which I simply could not google.
BTW: I have a chatgpt conversation about the issue, which I won't publish in this too public forum, but that I don't mind sharing with individuals.

Regards
Peter