On Fri, Mar 22, 2013 at 8:40 AM, Iry Witham iry.wit...@jax.org wrote:
I am working on setting up a Cloudman instance for a course we are
preparing. I have used the link provided: http://usegalaxy.org/cloudlaunch
and can get an instance running. The issues are as follow:
1. I need to add tools and have not found a clear method in doing so
utilizing the galaxy toolshed.
Until the new release of Cloudman there are a few manual steps you
need to take if you want to get an instance started that's compatible
with the toolshed interface.
Starting with a fresh Galaxy type instance, launched via
usegalaxy.org/cloudlaunch:
Update Galaxy. You can do this through the admin panel
(your_instance/cloud/admin), or by ssh'ing in and pulling. The cloud
instances previously tracked galaxy-central, though you'll probably
want to pull from -dist now.
Depending on how far you update (current -central tip for sure) this
may not merge all changes to run.sh cleanly. You'll need to ssh in
and add the additional .sample files. You should be able to simply
`hg diff run.sh` and then edit the file to add the missing samples.
While ssh'ed in, edit universe_wsgi.ini to both add yourself as an
admin user and configure the tool_config_file entry to look like this:
tool_config_file = tool_conf.xml,shed_tool_conf.xml
Now back in the cloudman admin interface, you'll want to find the
galaxy service and 'update db', and then 'restart' galaxy, potentially
having to 'restart' two times if tool migrations were included in your
update (these throw an exception at launch and exit, printing stuff to
the log).
After that, unless I've forgotten something, you should be good to go.
2. Once I have the instance setup and shut it down I cannot restart it
cleanly. On restart it appears that nginx fails to launch since port 80 is
not listening. I have not found a way to start it manually.
Stopping and starting an instance via the AWS console isn't supported,
it'll definitely break things. Instead, just use the terminate
cluster functionality in the primary cloudman interface. By default
all of the galaxy data (/mnt/galaxyData) is persisted. If you have
made modifications to the galaxyTools volume (by either manually
installing tools or using the toolshed to do so) you'll need to click
persist galaxyTools in the cloudman admin interface -- this is just
once, after modifications are done or tools are installed as you want
them. Of course, if you never want to use a cluster again and want to
clean up all saved data (both S3 and EBS) just check Delete
Permanently in cluster shutdown.
Good luck, and let me know if I've left anything out of my
instructions above and I'll do what I can to help!
Dannon
___
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/
