[galaxy-dev] job_conf.xml questions

1 Aug 2013

      Hi Ido Tamir,
I have the same question too. I set "megablast" as "long_jobs" for the test,when I cancel my megablast job in Galaxy history before it finishs,it still work in my background.The drmaa_external_killer.py seems doesn't work.I don't know where is the problem,and hope someone can help me .
a) In my advanced example I have:

        <destination id="long_jobs" runner="pbs" tags="cluster">
            <param id="Resource_List">walltime=72:00:00,nodes=2:ppn=8,mem=16G</param>
            <param id="galaxy_external_runjob_script">scripts/drmaa_external_runner.py</param>
            <param id="galaxy_external_killjob_script">scripts/drmaa_external_killer.py</param>
            <param id="galaxy_external_chown_script">scripts/external_chown_script.py</param>
        </destination>
    <tools>
        <tool id="ucsc_tablebrowser" destination="regularjobs"/>
        <tool id="megablast_wrapper" destination="long_jobs"/>
    </tools>

b)In my background,I get the "qstat" result :
Job id                    Name             User            Time Use S Queue
------------------------- ---------------- --------------- -------- - -----
46.server                  ...n@genomics.cn galaxy          00:00:00 R batch 

shenwiyn

From: Ido Tamir
Date: 2013-07-23 21:53
To: galaxy-dev@lists.bx.psu.edu
Subject: [galaxy-dev] job_conf.xml questions
Hi,
I work myself through the job_conf.xml and have a question:

a)
In your advanced example you have:

<destination id="remote_cluster" runner="drmaa" tags="longjobs"/>
 <destination id="real_user_cluster" runner="drmaa">
            <!-- TODO: The real user options should maybe not be considered runner params. -->
            <param id="galaxy_external_runjob_script">scripts/drmaa_external_runner.py</param>
            <param id="galaxy_external_killjob_script">scripts/drmaa_external_killer.py</param>
            <param id="galaxy_external_chown_script">scripts/external_chown_script.py</param>
 </destination>

Does this mean that remote_cluster jobs can not be killed unless I add the 3 scripts to this destination?
Or does this have to do with the "run jobs as real user" that I don't need currently. I am using the galaxy user for all jobs for the moment?

To phrase this differently: Do I need these three scripts?

b)
Can I set a wall time limit for only for my local runners?

like:
<destination id="local" runner="local"/>
   <param id="Resource_List">walltime=72:00:00</param>
</destination>

When I set it in limits, will it work for local runners and then I have to override it in 
every destination?

thank you very much,
ido

___________________________________________________________
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:
  http://lists.bx.psu.edu/

To search Galaxy mailing lists use the unified search at:
  http://galaxyproject.org/search/mailinglists/