rnaseq ARM Module updates #7101
Conversation
edmundmiller
force-pushed
the
enable-wave-failures
branch
from
f362c95 to
14d2324
Compare
edmundmiller
force-pushed
the
enable-wave-failures
branch
from
14d2324 to
51e4760
Compare
| @@ -52,12 +52,12 @@ jobs: | |||
| # if: github.repository == 'nf-core/modules' | |||
| if: ${{ needs.generate-matrix.outputs.conda-matrix != '[]' }} | |||
| needs: generate-matrix | |||
| name: Build ${{ matrix.profile }} ${{ matrix.platform }} Container | |||
| name: 🌊 ${{ matrix.files }} ${{ matrix.profile }} ${{ matrix.platform }} | |||
There was a problem hiding this comment.
🙅🏻 makes it difficult for newcomers to understand what is happening
There was a problem hiding this comment.
Yeah, I can agree with that.
I was trying to clean up the names because it's also hard for newcomers and me to figure out what's happening when the name is cut off 🙃
Open to suggestions 🤷🏻♂️
| runs-on: ubuntu-latest | ||
| timeout-minutes: 60 | ||
| strategy: | ||
| fail-fast: false | ||
| max-parallel: 4 | ||
| # max-parallel: 2 |
There was a problem hiding this comment.
why not 4? we have cores, so why not use them?
There was a problem hiding this comment.
- I turned it off for the purpose of running them all quickly
- I don't know exactly how that will translate to AWS ec2 runners. A lot of these jobs will just ping wave and be done. So I didn't want to spin up 20 runners because it's getting ready for a lot of runs and we're going to need a lot of capacity, and then it's 20 of the same wave runs because it was GATK.
There was a problem hiding this comment.
can we make this action fail gracefully for now, meaning that it exits with code 0 in any way? makes the launch of this less disruptive.
There was a problem hiding this comment.
I agree, we don't want all the CI to start failing. Maybe we could test locally for inspecting which containers are failing? and make this action always pass
There was a problem hiding this comment.
We've been doing that; this was @ewels' ask because he'd like a list of what modules are failing 🤷🏻♂️
Let's do a Slack hook or do that in an alternative script if we still want the wave jobs to pass.
There was a problem hiding this comment.
Are we adding successful builds to meta.yml? That'd be nice. Then we could just not add to meta.yml on fail but not trigger a CI failure for now.
A slack hook on build fail would be nice if easy 🤔
…ules - Bump bracken version from 2.9 to 3.0 - Update samtools version from 1.16.1 to 1.21 - Upgrade star version from 2.7.10a to 2.7.11b in both rsem modules
|
Let's strip out all changes unrelated to the new Wave CI |
edmundmiller
changed the title
Updates all of the versions in the rnaseq modules to ensure they have ARM conda builds.