Heyy! Anybody else that is already using the lates...
# pipeline-maintainers
e
Heyy! Anybody else that is already using the latest Nextflow version along with nf-core tools 3.2.0+ and still has a similar problem?: https://nfcore.slack.com/archives/C043FMKUNLB/p1747384407394789 I am kinda lost here
j
Nope, I've not seen it
e
Ooh, someone using mag saw it though: https://github.com/nextflow-io/nextflow/issues/6090#issuecomment-2888088575 That's hopeful for me xD
j
Weird
a
Could this issue be related to this other recent announcement? https://nfcore.slack.com/archives/CE6P95170/p1747302440656339
e
Could be, but couldn't be fixed by upgrading to nf-core tools 3.2.0
a
Havr you tested with older nf-core tools / nextflow versions?
e
Looking into the .command.run file, the .command.trace file is getting populated by the
nxf_mem_watch
function (lines 105-112) and the error comes from within the
nxf_trace_linux
function (lines 182-192). Not sure if this helps.
o
Do you have the whole traceback of the error? If I understand well, you get this error for all processes in you pipeline? Or is it only all instances of a specific process?
e
It isn't an error that gets printed on the console. Just a line that .command.run prints on .command.err and .command.log, and then the .command.trace is wrong. I've posted some more information on the GitHub issue above
👍 1
^it happens for all processes
👍 1
o
Actually the same thing is happening for me. I've created last week a new pipeline (for personal purposes) using nf-core/tools v3.2.1, and in all workdirs, the .command.run file shows the same error. However, nextflow runs / nf-tests are still successful
e
Yep! If you try -with-tower, you will only see Job Duration barplots in the Metrics tab, whereas all the rest (cpu, memory, I/O) will be empty
👍 1
a
Perhaps it could also be worth linking this GitHub issue in #CE5LG7WMB
e
I think I've already spammed quite a few channels with it 😛
😂 1
a
The more people are bothered by it, the better 😄 It seems to be an issue that potentially affects plenty of pipelines
👍 1
j
Did you try playing with the pipefail stuff in your pipeline @Evangelos Karatzas?
e
Nope, don't even know what that is
j
See what Phil said in the announcemnet
I think it's the same issue
e
Ah that one, it must be related yes
The 3.2.1 template was supposed to fix this, and I checked the code and it was agreeing with the lines that were supposed to change
j
I mean play around with it, don't rely on the template 😬
🥲 1