[CLI] Job sent to cluster never gets back

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

[CLI] Job sent to cluster never gets back

Estelle Ancelet
Hello Galaxy community,

We are trying to connect a Galaxy web server to a remote cluster by using the "CLI" approach.
We do partially succeed : jobs are submitted to the cluster but seem not to be monitored. That is, jobs are executed on the cluster but Galaxy doesn’t update the status in the history (stays grey) and never gets back the outputs.

The Galaxy folder on the cluster is mounted to the Galaxy Server using sshfs :
sshfs [hidden email]:galaxy galaxy
Galaxy (v18.05) is launched by a “sh run.sh” in the galaxy folder, from the Galaxy server side.

Here is an extract of our job_conf.xml:
        <plugin id="local" type="runner" load="galaxy.jobs.runners.local:LocalJobRunner" workers="4"/>
        <plugin id="cli" type="runner" load="galaxy.jobs.runners.cli:ShellJobRunner" workers="4"/>
    <destinations default="ssh_slurm">
        <destination id="local" runner="local"/>
        <destination id="ssh_slurm" runner="cli">
            <param id="shell_plugin">SecureShell</param>
            <param id="job_plugin">Slurm</param>
            <param id="shell_username">galaxyuser</param>
            <param id="shell_hostname">cluster.fr</param>
            <param id="job_time">2:00:00</param>
            <param id="job_ncpus">4</param>
            <param id="job_partition">defq</param>
        <tool id="runtask" destination="ssh_slurm" />

Did we miss something in the configuration? Or do we have to write additional code?

Sincere thanks,
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

To search Galaxy mailing lists use the unified search at: