Hi all,
I have jobs that need to be queued up and run in a single thread
(the labor intensive part of the job uses a library that is not thread
safe)
Right now, I have a single worker. The thread pool_size = 1 but jobs
are put in using thread_pool.defer so that backgroundrb takes care of
the queue for me. This works fine except that the queue gets larger
than I would like. I'd like to have multiple identical (single
threaded, own Rais environment) workers sharing a queue and processing
the jobs.
I could start up several workers with different worker_keys and
distribute jobs on the client side, but this seems a little messy. Are
there any better options?
I'm using the latest backgroundrb/packet from git and memcached for
the result cache.
Casey
_______________________________________________
Backgroundrb-devel mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/backgroundrb-devel