I've read up on a few approaches to this, and wondering if my (brute)
approach below will work.  Suggestions welcome!

The application servers will all run searchd, and will be creating
local delta's.  It will also be running off the main index file, which
will be created on our database slave nightly, scp'ed over, where the
searchd is then hup'ed.

My only issues with this approach are:

1.  I'll could have application server count + (N) misses on the
deltas, as they won't be consistent across servers.
2.  I have to scp the main indexes to the all the application
machines.

Now, there are a few other distributed methods I've seen, which use
agents to communicate to all the other servers, pulling from the
delta's, then returning a diff'ed result set.  My only issue with this
approach is that we have about 30+ application servers, and many get
added and removed frequently.

Having to manually update the agent addresses in the config is out of
the question, so, are there any auto-detection methods available?

I'm probably missing something, so if anybody has any input/advice, it
would be greatly appreciated :)
Thanks!
==
Dylan

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Thinking Sphinx" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/thinking-sphinx?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to