Thanks Nate. I have tried that but it seems to run into problems with the expected output. This may be simply the aforementioned python dependency issues, but I'll have to investigate more in the coming week when I have time. Just to get an idea of what I would like, I mainly want the capability to run tests on (for instance) a new Galaxy deployment on an already-established cluster or other similar conditions where tools may not be configured correctly (e.g. missing secondary dependencies) or subtle versioning issues exist (e.g. wrong version of cufflinks). I'm thinking this could eventually lead to a mechanism where one optionally generates a tools_conf file that comments out or removes tool configurations not working properly, as well as generates a report of the issues found (as run_functional_tests already does). chris On Jan 20, 2012, at 8:54 AM, Nate Coraor wrote:
On Jan 19, 2012, at 12:57 AM, Fields, Christopher J wrote:
Just to add for the list, nate mentioned on IRC that running tests on an external server likely doesn't work at the moment but that ./scripts/functional_tests.py may have something to work with via environment variables.
The tool config can be specified in:
$GALAXY_TEST_TOOL_CONF
The config file is a bit trickier since many Galaxy config options can be overridden by environment variables or are set to something other than the normal default right in functional_tests.py. You could change this line (147):
global_conf = { '__file__' : 'universe_wsgi.ini.sample' }
To change the config default, however.
--nate
chris
On Jan 18, 2012, at 8:03 PM, Fields, Christopher J wrote:
Greg,
That seems to work mainly for running tests from the sample universe/tool config files, so it uses the local job runner. What I would like to do is run tests using the actual Galaxy server configuration (which might use PBS/DRMAA and a cluster). The method I mention from that page seemed to be for that purpose, but again it seems possibly out-of-date...
chris
On Jan 18, 2012, at 4:20 PM, Greg Von Kuster wrote:
Hello Chris,
You should be able to run functional tests from the Galaxy install directory using the following from the command line.
%sh run_functional_tests.sh
There are several options you can use as well.
Greg
On Jan 18, 2012, at 4:38 PM, Fields, Christopher J wrote:
I have been following the basic protocol for running remote tests on a Galaxy server (in this case, our local one) using:
http://wiki.g2.bx.psu.edu/Admin/Running%20Tests
From reading this, I assume the tests are to be run from the galaxy root directory, like so:
GALAXY_TEST_EXTERNAL=1 GALAXY_TEST_HOST=my.local.galaxy.url GALAXY_TEST_PORT=8080 \ ./nosetests.sh
However, no 'nosetests.sh' script exists there. Should we be using ./scripts/nosetests.py? ./scripts/functional_tests.py?
Any clarification on this would be very helpful, we're just searching for a way to run tests in as automated a way as possible using the server configuration.
chris
Chris Fields Senior Research Scientist National Center for Supercomputing Applications Institute for Genomic Biology University of Illinois at Urbana-Champaign
___________________________________________________________ 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:
Greg Von Kuster Galaxy Development Team greg@bx.psu.edu
___________________________________________________________ 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:
___________________________________________________________ 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: