On Wed, Apr 9, 2014 at 9:55 AM, Alexandre Defelicibus <adefelicibus@gmail.com> wrote:
Hi all,
I have some doubts about how to configure a LWR server on Linux. I have read the LWR documentation, however how to configure the LWR is not clear to me.
This is my scenario:
On my Galaxy server I have some tools that I would like to send their jobs to the LWR server. These tools are at the same folder (xml and py).
The LWR server is installed on Linux, and it is accessible through the URL.
This is my job_conf.xml:
<?xml version="1.0"?> <job_conf> <plugins workers="4"> <plugin id="lwr" type="runner" load="galaxy.jobs.runners.lwr:LwrJobRunner"> </plugin> </plugins> <handlers default="handlers"> <handler id="handler0" tags="handlers"/> </handlers> <destinations default="local"> <destination id="local" runner="local"/> <destination id="lwr" runner="lwr"> <param id="url">http://200.144.255.42:8005/</param> </destination> </destinations> <tools> <tool id="2pg" destination="lwr"/> </tools> <limits> <limit type="registered_user_concurrent_jobs">2</limit> <limit type="unregistered_user_concurrent_jobs">1</limit> <limit type="job_walltime">24:00:00</limit> <limit type="concurrent_jobs" id="local">1</limit> </limits> </job_conf>
When I execute the tool, I get this error:
ValueError: No JSON object could be decoded
At the LWR server, I use a nginx server, and I see on the log file that the server receive this URL: GET /setup?tool_id=2pg&job_id=24&tool_version=1.0.0
Cool I have never setup LWR behind nginx but it should certainly work.
So, I concluded that the Galaxy server is accessing the LWR server and creating a job, but something wrong happens on the LWR side, but I couldn't realize what is missing.
How are you running the LWR? There should be a file called paster.log in the LWR root directory or in the standard out of the terminal session you started the LWR using if it did not start as a daemon. That should have a lot of relevant details. Basically there should be a bunch of lines in Galaxy's log and the LWR log corresponding to the attempted submission that will hopefully provide some details related to the error.
So, I have some questions to ask:
- Do I need to have a Galaxy instance running at the same server of LWR?
No this definitely should not be needed.
- Are there other variables that I should have set on Galaxy or on LWR server?
Not really....
- I know it is better if I use the secure LWR, and I will do, but is there something wrong on my job_conf.xml file?
Nothing jumps out at me. The logs will make it clear if there is though I think. The LWR is distributed with a test script to exercise the server independent of Galaxy - you should target it at your nginx proxy to determine if the problem is with above the proxy (i.e. in Galaxy's configuration or the tool itself) or below it (either in the LWR, nginx, etc...).
From the LWR root you should be able to just run:
python run_client_tests.py --url http://200.144.255.42:8005/ You could also clone down the LWR code on your Galaxy host and run this to determine if there is some firewall problem for instance. Hope this helps and thanks for your interest in the LWR! -John
I hope I have shown all the needed details about my installation.
Best regards,
Alexandre
-- Alexandre Defelicibus Mestrando em Bioengenharia Programa de Pós-Graduação em Bioengenharia Universidade de São Paulo - USP
___________________________________________________________ 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/