training
training copied to clipboard
nextflow-io
Pipeline mentioned in section `3.10` stopped working
The following pipeline stopped working:
nextflow run rnaseq-nf -with-docker -with-report -with-trace -with-timeline -with-dag dag.png
It works if one specifies the dev revision with -r dev, but this is a very old one and shouldn't be used. By investigating the errors, it seems to be the way modules are called in the include instruction.
Which container version are you running?. Ah, maybe the old container does not work because the code has changed in rnaseq-nf. The new container works fine (https://github.com/seqeralabs/nf-training-public/pull/114), so we should switch asap.
@chriswyatt1 , maybe we should make use of the tags here to maintain reproducibility?
The workaround of using -r dev worked, but the dev branch is really quite behind (~2 years).
@abhi18av how do you mean? As in we should publish the current branch of the training with the older container as V1, and make the new container version V2?
Exactly!
Since we rely heavily on the rnaseq-nf pipeline during the training it's better to make use of the tags so that we don't run across surprises within the training material and the latest master can continue moving forward to accommodate new developments in NF runtime.