search

bgruening / docker-galaxy

:whale::bar_chart::books: Docker Images tracking the stable Galaxy releases.
http://bgruening.github.io/docker-galaxy
MIT License
226 stars 134 forks source link

"NONUSE=slurmctld,slurmd" causes jobs not be executable #490

Closed FelixErnst closed 5 years ago

FelixErnst commented 5 years ago

When create a container using docker run -d -p 8080:80 -e "NONUSE=slurmctld,slurmd" -v /home/user/galaxy_storage/:/export/ bgruening/galaxy-stable, the galaxy instance cannot execute jobs.

Loading a reference genome via Admin-> local data (sacCer3 for example) does not get executed. The error is "Unable to run this job due to a cluster error, please retry it later". The same error occurs, when uploading files

Is this option a non viable configuration? Since I had the same problems as described in #489, I thought maybe the type of worker had an influence on this.

Any help or advice is much appreciated

bgruening commented 5 years ago

Hi Felix,

we need to use SLURM in a standard setup. This NONUSE option is only useful if you use an outside SLURM cluster. So for example if Galaxy should submit jobs to your HPC. In this case SLURM inside the container does not make sense.

Does this answer your Question? If not please reopen!

Thanks for using Galaxy, Bjoern

FelixErnst commented 5 years ago

Hi Björn, thanks for the clarification. It does answer my question.

Slightly off-topic: since opening the issue a few hours ago, I am wondering how can I adjust the number of workers per machine? I didn't find anything in the docs. Did I overlook something?

bgruening commented 5 years ago

You can control this with some magic ENV variables: Please see here: https://github.com/bgruening/docker-galaxy-stable#toc