Hi,
I have some tools that run really quickly without using any kind of cluster. I would prefer not to run these tools on a cluster, as the overhead of submitting these jobs makes them take much longer than they otherwise would. I have other tools that are computationally intensive and need to be run on a cluster. I would like to expose all these tools in the same Galaxy instance, but have some tools run on the cluster and others not.
Is this possible?
Thanks, Dan
Hey Dan,
Sure, you can configure per-tool job runners. This wiki page ( http://wiki.galaxyproject.org/Admin/Config/Jobs) should get you started, but let me know if you run into any troubl
On Tue, Apr 30, 2013 at 4:47 PM, Dan Tenenbaum dtenenba@fhcrc.org wrote:
Hi,
I have some tools that run really quickly without using any kind of cluster. I would prefer not to run these tools on a cluster, as the overhead of submitting these jobs makes them take much longer than they otherwise would. I have other tools that are computationally intensive and need to be run on a cluster. I would like to expose all these tools in the same Galaxy instance, but have some tools run on the cluster and others not.
Is this possible?
Thanks, Dan ___________________________________________________________ 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/
On Tue, Apr 30, 2013 at 2:04 PM, Dannon Baker dannon.baker@gmail.com wrote:
Hey Dan,
Sure, you can configure per-tool job runners. This wiki page (http://wiki.galaxyproject.org/Admin/Config/Jobs) should get you started, but let me know if you run into any trouble.
Thanks!
Following up on this.
As I mentioned before, this Galaxy instance was already configured to use the cluster by default. I did not set up the cluster or the Galaxy instance. This particular instance did not use the job_conf.xml file. But I updated it and set up a job_conf.xml and I was able to set it up so that some tools would be run on the local machine.
However, in trying to get it set up so some jobs would run on the cluster (which is what happened by default before I came along) I ran into problems. Previously, it looked like the cluster stuff was configured in universe_wsgi.ini, and I guessed that these were the relevant lines:
start_job_runners = drmaa default_cluster_job_runner = drmaa://-t 12:00 -A noaccount/
So I tried to move that setup to job_conf.xml:
?xml version="1.0"?> <!-- A sample job config that explicitly configures job running the way it is configured by default (if there is no explicit config). --> <job_conf> <plugins> <plugin id="local" type="runner" load="galaxy.jobs.runners.local:LocalJobRunner" workers="4"/> <plugin id="drmaa" type="runner" load="galaxy.jobs.runners.drmaa:DRMAAJobRunner"/> </plugins> <handlers> <handler id="main"/> </handlers> <destinations default="local"> <destination id="local" runner="local"/> <destination id="cluster" runner="drmaa"> <param id="nativeSpecification">-t 12:00 -A noaccount/</param> </destination> </destinations> <tools> <!-- Tools can be configured to use specific destinations or handlers, identified by either the "id" or "tags" attribute. If assigned to a tag, a handler or destination that matches that tag will be chosen at random. --> <tool id="mytesttool" destination="cluster"/> </tools> </job_conf>
When I run 'my test tool', however, it just hangs.
Any tips? Thanks, Dan
On Tue, Apr 30, 2013 at 4:47 PM, Dan Tenenbaum dtenenba@fhcrc.org wrote:
Hi,
I have some tools that run really quickly without using any kind of cluster. I would prefer not to run these tools on a cluster, as the overhead of submitting these jobs makes them take much longer than they otherwise would. I have other tools that are computationally intensive and need to be run on a cluster. I would like to expose all these tools in the same Galaxy instance, but have some tools run on the cluster and others not.
Is this possible?
Thanks, Dan ___________________________________________________________ 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/
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/
On Tue, May 7, 2013 at 10:16 AM, Dan Tenenbaum dtenenba@fhcrc.org wrote:
On Tue, Apr 30, 2013 at 2:04 PM, Dannon Baker dannon.baker@gmail.com wrote:
Hey Dan,
Sure, you can configure per-tool job runners. This wiki page (http://wiki.galaxyproject.org/Admin/Config/Jobs) should get you started, but let me know if you run into any trouble.
Thanks!
Following up on this.
As I mentioned before, this Galaxy instance was already configured to use the cluster by default. I did not set up the cluster or the Galaxy instance. This particular instance did not use the job_conf.xml file. But I updated it and set up a job_conf.xml and I was able to set it up so that some tools would be run on the local machine.
However, in trying to get it set up so some jobs would run on the cluster (which is what happened by default before I came along) I ran into problems. Previously, it looked like the cluster stuff was configured in universe_wsgi.ini, and I guessed that these were the relevant lines:
start_job_runners = drmaa default_cluster_job_runner = drmaa://-t 12:00 -A noaccount/
So I tried to move that setup to job_conf.xml:
?xml version="1.0"?>
<!-- A sample job config that explicitly configures job running the way it is configured by default (if there is no explicit config). -->
<job_conf> <plugins> <plugin id="local" type="runner" load="galaxy.jobs.runners.local:LocalJobRunner" workers="4"/> <plugin id="drmaa" type="runner" load="galaxy.jobs.runners.drmaa:DRMAAJobRunner"/> </plugins> <handlers> <handler id="main"/> </handlers> <destinations default="local"> <destination id="local" runner="local"/> <destination id="cluster" runner="drmaa"> <param id="nativeSpecification">-t 12:00 -A noaccount/</param> </destination> </destinations> <tools> <!-- Tools can be configured to use specific destinations or handlers, identified by either the "id" or "tags" attribute. If assigned to a tag, a handler or destination that matches that tag will be chosen at random. --> <tool id="mytesttool" destination="cluster"/> </tools> </job_conf>
When I run 'my test tool', however, it just hangs.
Any tips?
To answer my own question...I think I have gotten this working, by using the drmaa snippet from the job_conf.xml.sample_advanced file in my galaxy distro:
<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>
If I set up my test tool to run with this destination, it works.
Thanks, Dan
Thanks, Dan
On Tue, Apr 30, 2013 at 4:47 PM, Dan Tenenbaum dtenenba@fhcrc.org wrote:
Hi,
I have some tools that run really quickly without using any kind of cluster. I would prefer not to run these tools on a cluster, as the overhead of submitting these jobs makes them take much longer than they otherwise would. I have other tools that are computationally intensive and need to be run on a cluster. I would like to expose all these tools in the same Galaxy instance, but have some tools run on the cluster and others not.
Is this possible?
Thanks, Dan ___________________________________________________________ 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/
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/
galaxy-dev@lists.galaxyproject.org