https://nf-co.re logo
Join Slack
Powered by
:alert: AWS megatest GHA auto-submission currentl...
# pipeline-maintainers
j
alert AWS megatest GHA auto-submission currently INOP alert Please note that auto-submission of megatests to Seqera Platform on 
master/main
-> 
dev
Prs is currently not working! Please ping a <!subteam^S016E36N7TQ> for them to execute this test manually for you once you have your release reviews accepted More details 👇 https://nfcore.slack.com/archives/CE6SDBX2A/p1755801259022109?thread_ts=1755739921.163179&amp;channel=CE6SDBX2A&amp;message_ts=1755801259.022109
r
the pipeline maintainers should have launch permissions to run their pipelines from platform. If you don't have it someone from the core team can add you
j
Can they make new pipleines to launchapd though?
a
^ I don't think so? I can't do it. Just did a patch release for denovotranscript. Can someone please help me launch it?
e
a
no, I'm seeing an access denied message
e
Okay bumped you from launch to maintain, can you test again?
a
yes, that works. Thank you!
🎉 1
but I'm getting a API rate limit exceeded message when I click launch
p
Do you have your GitHub credentials in your Seqera platform user credentials?
n
Hey, there is still the issue with the #CHN5BV5DW release. Afaik megatests have already been working in earlier releases. I got launch access to nf-core/AWSmegatests a couple of days ago, but I cannot find anything related to #CHN5BV5DW. Am I missing something? Otherwise could I also get elevated permissions?
e
Hey @Nico Trummer, could you link me to the scrnaseq issue you're referring to?
n
e
Gotcha, so seeing the same issue in the actions. Are you able to launch in platform?
n
I got launch access to nf-core/AWSmegatests a couple of days ago, but I cannot find anything related to #CHN5BV5DW. Am I missing something? Otherwise could I also get elevated permissions?
I think this describes the situation pretty well
e
Sorry, let me clarify. You're not able to launch without a configuration?
n
No this button does not exist for me
👍🏻 1
e
Okay, you should have permissions now. It's not going to fix the GitHub actions, but you can run the test full profile.
n
Worked, thanks. Should I use the "Run without configuration" now or add an #CHN5BV5DW config?
e
Whichever you prefer! We'll need to move the results to the correct bucket/prefix for them to display on the site.
d
That isnt fixed yet, right? I seem to be not able to start the AWS tests from github. I can access https://cloud.seqera.io/orgs/nf-core/workspaces/AWSmegatests/launchpad but I have no idea to start from there. Just wanted to make sure my new release runs on AWS... will release now anyway.
r
do you want me to talk you through it? Essentially, you can click on "Launch without configuration", select 
aws_ireland_fusionv2_nvme_cpu
and then fill out the rest. The tricky part is getting the 
outdir
right which should use the git sha of the release. If you are just running against dev, you can use 
results-dev
d
Thanks Rike! However, I do not see that line at the bottom:
n
Hey, I have been trying to get the megatests for #CHN5BV5DW running, but I keep getting an error that I don't know how to fix. Could someone with more experience check this run and maybe give me a pointer?
j
I've just tried switching to a nofusion run and see if the same thing happens
r
I don't think this is a fusion issue. We have seen this before when something was wrong with tags assigned in the CE
j
It might not be, but it's my usual first attempt 😉
The non-fusion run is going here @Nico Trummer in the meantime https://cloud.seqera.io/orgs/nf-core/workspaces/AWSmegatests/watch/1oQx5A7fm5wT8r/v2/tasks
n
Thanks for taking this on!
j
This will be the extent of my ability to help debugging though 😅 , so I'll leave it with Rike after this
r
yep same error, think its either sometihng in the publishing logic. I resubmitted your original run here without tagging the output: https://cloud.seqera.io/orgs/nf-core/workspaces/AWSmegatests/watch/WIITCw1jCDWi5/v2/logs
it's interesting that this onlye ffects some workflows.
Copy code
Error executing process > 'NFCORE_SCRNASEQ:SCRNASEQ:SIMPLEAF:SIMPLEAF_INDEX (genome.fa)'

Caused by:
  Job attempt duration exceeded timeout
Normally all the pipeline maintainers should have launch access to platform so you should be able to this yourself. If you do not please react to this message so I can add you.
seqera 1
m
We fixed the faulty aws launch action in the new 3.4.1 template. if you don't want to merge in the full template update, but fix your current versions, apply the following changes https://github.com/nf-core/tools/pull/3827/files
🎉 4
🦸 1
n
Hey, I encountered some more problems after performing the template update in #CHN5BV5DW : • An 'Access denied' in the runs launched via the GitHub actions on Seqera Platform • 'Conflict: an artifact with this name already exists on the workflow run' directly in the GitHub action
m
Yes, @Maxime Garcia ran into the same permission error on aws. I will investigate today a bit more
👀 1
m
Oh yeah what I mentioned yesterday in #C069B2Z5P1R
n
The second Problem I mentioned above was fixed by @Gregor Sturm via this PR The 'Access denied' is still there
2 Views
Open in Slack
PreviousNext
Powered by