We run a local galaxy instance with multiple galaxy web servers and a job runner. Per the wiki suggestion we have two config files: universe_wsgi.webapp.ini and universe_wsgi.runner.ini. I've noticed that we still need universe_wsgi.ini or else galaxy will not start Instead of maintaining three config files (which are nearly identical) it would be really helpful if we could just have a single config file. As far as I can tell, the only thing preventing me from doing this is the "enable_job_running" option. If this were a server parameter, then I could have a single config file. Something like this: [server:runner] use = egg:Paste#http port = 8079 host = 0.0.0.0 use_threadpool = true threadpool_workers = 3 enable_job_running = True [server:web0] use = egg:Paste#http port = 8080 host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False [server:web1] use = egg:Paste#http port = 8081 host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False ... [server:webN] use = egg:Paste#http port = 808X host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False With a single file I wouldn't have to figure out which of our config file(s) a new setting needs to go into, I can just make the change once
Glen Beane wrote:
We run a local galaxy instance with multiple galaxy web servers and a job runner. Per the wiki suggestion we have two config files: universe_wsgi.webapp.ini and universe_wsgi.runner.ini. I've noticed that we still need universe_wsgi.ini or else galaxy will not start
Hi Glen, Yes, unfortunately this is a known issue we haven't had a chance to correct yet: https://bitbucket.org/galaxy/galaxy-central/issue/358/the-config-file-is-har...
Instead of maintaining three config files (which are nearly identical) it would be really helpful if we could just have a single config file. As far as I can tell, the only thing preventing me from doing this is the "enable_job_running" option. If this were a server parameter, then I could have a single config file. Something like this:
[server:runner] use = egg:Paste#http port = 8079 host = 0.0.0.0 use_threadpool = true threadpool_workers = 3 enable_job_running = True
[server:web0] use = egg:Paste#http port = 8080 host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False
[server:web1] use = egg:Paste#http port = 8081 host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False
...
[server:webN] use = egg:Paste#http port = 808X host = 0.0.0.0 use_threadpool = true threadpool_workers = 7 enable_job_running = False
With a single file I wouldn't have to figure out which of our config file(s) a new setting needs to go into, I can just make the change once
I agree this would be nice. It's something I've put a bit of thought into but haven't had a chance to work on yet. The blocking factor is that a server has no way of knowing which [server:] block it was started from without some dirty hacks (i.e. the section name or any other identifier is not actually passed into the application when the factory is built). --nate
_______________________________________________ To manage your subscriptions to this and other Galaxy lists, please use the interface at:
participants (2)
-
Glen Beane
-
Nate Coraor