Changing the default job manager is not interchangeable in an automated fashion. SGE is currently the main job manager that Galaxy has been configured to be used with. Slurm will replace SGE in the next Galaxy CloudMan release. HTCondor has been added as an experimental module but due to the lack of interest from the community (ie, no interest until now), it hasn't seen any development and I'm really not even sure if it's working as expected any more. Relevant documentation is available here https://wiki.galaxyproject.org/CloudMan/HTCondor so if you can try this out first and see what happens; we can then go from there.
Having said that, if you install any of the desired job managers manually, Galaxy can be configured to use that one by setting appropriate configuration option in */mnt/galaxy/galaxy-app/universe_wsgi.ini*. See this page for more details on configuring Galaxy: https://wiki.galaxyproject.org/Admin/Config/Performance/Cluster
Regarding the use of an S3 bucket as a data source - the intent (largely mandated by the technical characteristics of an object store) is for this to be used as a source of data rather than the default data store for Galaxy. With that, it is advisable to use it as a source location for importing shared datasets and ultimately storing the resulting ones. What this means from within Galaxy is that you could create a data library with the mounted S3 bucket being a data source and then utilize it for the analyses. Take a look at slides #25-28 in this presentation on how to do this: http://www.slideshare.net/afgane/data-analysis-with-galaxy-on-the-cloud If you would still like to use the S3 bucket as a comprehensive data store, this can be done by setting *file_path* in */mnt/galaxy/galaxy-app/universe_wsgi.ini* to the desired path.
Hope this clarifies some questions. Let us know what you decide to do and how things develop, Enis
On Tue, Oct 14, 2014 at 6:35 AM, Abdularahman Azab azab@ifi.uio.no wrote:
Dear all,
I've installed a galaxy cloud cluster using usergalaxy.org/cloudlaunch
It runs fine, but I need to know how to let the cluster run on THCondor or slurm instead of SGE. I can see that Condor is already installed on the galaxy-master node since the condor_status command is there. But how can I switch the default galaxy job-runner between SGE, HTCondor, and slurm
Also regarding the mounting option of a e.g. s3 bucket on the master node. It works fine and it can be accessed from the command line. But how to let galaxy see this mounted path for e.g, reading input files and storing resulted history elements in it?
Thank you,
Yours sincerely, Abdulrahman Azab
Head engineer, ELIXIR.NO / The Genomic HyperBrowser team Department of Informatics, University of Oslo, Boks 1072 Blindern, NO-0316 OSLO, Norway Email: azab@ifi.uio.no http://owa/abdulara@ifi.uio.no, Cell-phone: +47 46797339
Senior Lecturer in Computer Engineering Faculty of Engineering, University of Mansoura, 35516-Mansoura, Egypt Email: abdulrahman.azab@mans.edu.eg http://owa/abdulrahman.azab@mans.edu.eg
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/