[Backgroundrb-devel] Using the ip flag

Noah Horton noah at rapouts.com
Wed Mar 12 16:04:39 EDT 2008

Actually, my requirements are a bit different.  Sorry for being unclear.  Thanks for your help on this!

My setup is that I have a cluster of machines (lets call them A, B and C) that are all running clusters of mongrels (lets say 3 per machine, labeled A1, A2, A3, B1, B2, B3, C1, C2, C3).

I want to setup another machine, D, such that it runs nothing but BackgroundRB.  All of the mongrels on all the machines (A, B and C) would call into the single instance of BackgroundRB on D.  Is this possible?

On 3/11/08 6:21 PM, "hemant kumar" <gethemant at gmail.com> wrote:

Hi Noah,

Sorry for late reply, I have been terribly busy with my day job and now
catching up on mails.

On Sun, 2008-03-09 at 15:31 -0700, Noah Horton wrote:
> Hey All,
>   I am looking for an example of the proper config settings for
> background rb to run my rails app and my backgroundrb instance on
> separate machines (or more precisely, to have a rails cluster with one
> backgroundrb machine).  Does anyone have such examples?

Okay, so you need one BackgrounDRb server for each machine so as mongrel
clusters running on that machine has access to BackgrounDRb server
running on that machine?

If yes, you can simply use "" as ip, and your mongrels running on
that machine will only use BackgrounDRb server running on that machine.

Also, I will advise you to use MemCache cluster for storing
status/result objects so as even if next request goes to mongrels
running on another machine, you will have access to result objects from
another BackgrounDRb server. I dunno, if I am answering your questions,
but if you have doubt shoot back.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/backgroundrb-devel/attachments/20080312/373317fc/attachment.html 

More information about the Backgroundrb-devel mailing list