Question: Galaxy not starting correctly - pbs_python error?
1
gravatar for balaszone
24 months ago by
balaszone60
balaszone60 wrote:

Dear all,

I'm trying to use pbs to queue jobs in our cluster. So I modified my job_conf.xml file to reflect 'pbs' queuing. Here is my job_conf.xml

job_conf.xml:

<job_conf> <plugins> <plugin id="local" type="runner" load="galaxy.jobs.runners.local:LocalJobRunner" workers="4"/> <plugin id="pbs" type="runner" load="galaxy.jobs.runners.pbs:PBSJobRunner" worker="10"/> </plugins> <handlers> <handler id="main"/> </handlers> <destinations default="batch"> <destination id="batch" runner="pbs"/> <destination id="big_jobs" runner="pbs"> <param id="Resource_List">walltime=72:00:00</param> </destination> </destinations>

</job_conf>

When I start the galaxy server, I get an error below:

. . AssertionError: The Python pbs-python package is required to use this feature, please install it or correct the following error: ImportError /hpcdata/galaxy/.venv/lib/python2.7/site-packages/pbs/_pbs.so: undefined symbol: pbs_submit_hash_ext


Documentation on WIki suggests to modify eggs.ini file but running, ./scripts/check_eggs.py indicates that eggs have been replaced with wheels. So I'm not sure if that is an issue.

I have already set the path for $DRMAA_LIBRARY_PATH as /usr/lib64/libdrm.so.2 (again there is no libdrmaa.so as indicated in wiki)

Also, the pythons version is as below python --version Python 2.7.5

Please help!!

THAnks, Bala

queue jobs galaxy pbs • 680 views
ADD COMMENTlink modified 24 months ago by Jennifer Hillman Jackson25k • written 24 months ago by balaszone60

Can anybody help with the pbs_python installation error especially with this newer version of Galaxy. All the WIKI docs reflect instructions for older version and i doesnt help much.

I read that eggs have been replaced with wheels in this newer version and not sure how one would go about the pbs_python installation with this version.

any help?

thanks!

ADD REPLYlink written 24 months ago by balaszone60

It looks like this is due to an old version of libtorque. Try updating it.

ADD REPLYlink written 24 months ago by Devon Ryan1.9k

And how do I that? Can I update without root permissions? Tx

ADD REPLYlink written 24 months ago by balaszone60

You'd need to install a different version and ensure that that's what get linked against. Having said that, your life would be easier if you did this somewhere where you're the system administrator (or can use docker).

ADD REPLYlink written 24 months ago by Devon Ryan1.9k

Hi, Pardon my lack of knowledge in understanding this installation. Is libtorque part of some application? I have installed the latest version of Galaxy assuming everything should have been updated. Also, WIKI suggested this compilation :- LIBTORQUE_DIR=/path/to/libtorque python scripts/scramble.py -e pbs_python, But i dont see scramble.py in my scripts directory. I just dont seem to figure out this simple pbs queuing via galaxy local instance. :(

Any other help is really appreciated.

Thanks! Bala

ADD REPLYlink written 24 months ago by balaszone60

libtorque is part of torque (I assume it installs the library by default). You can find it with the find command.

ADD REPLYlink written 24 months ago by Devon Ryan1.9k

hi, thanks for your quick reply. I did find the location of the file and set the environment variable accordingly. However, now I'm getting a different error when starting galaxy.

DrmCommunicationException: code 2: unable to contact qmaster using port 6444 on host "master.cm.cluster"

Any thoughts?

THAnks again.

ADD REPLYlink written 24 months ago by balaszone60

Presumably there's some configuration issue somewhere (outside of Galaxy). Maybe your cluster admin knows.

ADD REPLYlink written 24 months ago by Devon Ryan1.9k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 126 users visited in the last hour