We have moved over to slack, get your invite here (!): http://nf-core-invite.herokuapp.com and visit us https://nfcore.slack.com
nf-core-bot on api-doc
Travis build: 816 (compare)
apeltzer on dev
Markdown parsing: Fix all the e… Add new markdown rules to templ… Apparently whitespace isn't eno… and 2 more (compare)
With apologies for the cross-posting from the Nextflow Gitter (hadn't actually twigged there was a separate nf-core channel):
I'm coming to the conclusion that running nf-core pipelines (at least out-of the box) on Kubernetes is a non-starter due to custom functions in the config file which don't play nice with k8s (according to previous statements by Paolo). Would anyone dispute that conclusion? Or am I missing an easy workaround? We're evaluating whether deploying the nf-core workflows this way is going to be feasible.
Specifically: if I run:
nextflow kuberun ebi-gene-expression-group/rnaseq -profile test,docker -v nextflow
(this is just the stock nf-core rna-seq workflow with a minor tweak to fix paths to assets), I get:
Unknown method `check_max` -- Did you mean?
check_log
Exception in thread "Task submitter" groovy.lang.MissingMethodException: No signature of method: Script_5ee9b41e.check_max() is applicable for argument types: (Integer, String) values: [2, cpus]
check_max
function that’s the problem, I guess you can just override the config files with a custom config that doesn’t use the function if you want though.
base.config
from the pipeline, remove all check_max
function calls and then use that config file with -c