-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Erik and Eric,

thank you both for the info. I'll probably be setting up a test instance
sometime in the near future so we can kick the tires, as it were.

Erik A. Dahl wrote:
> Tom,
> 
> Just a couple thoughts to add here...
> 
> On Jan 10, 2007, at 10:15 AM, Eric Newton wrote:
> 
>>> number of sites. There are a couple of things I can't seem to find a
>>> lot of information on in the FAQ or Admin Guide, though, specifically:
> 
> We continue to improve the docs each release. :)
> 
>>>
>>> 1) How well does Zenoss scale across large (10K+ nodes) environments
>>> that are spread across multiple sites? Do you have any performance
>>> numbers for things like parallel service or host checks? Each node
>>> would have multiple checks running against them for different services
>>> and/or processes.
> 
> When you say nodes do you mean a device (ie a piece of hardware running
> an OS with interfaces filesystems, etc) or individual monitors (like a
> single filesystem check)?  We scale the system by thinking of "devices"
> which may have many monitors each.
> 

Yes, node == device. We have a distributed computing cluster and that
nomenclature seems to slip out when I least expect it.

> Certainly the system is designed to scale to large numbers and to that
> end we can distribute our collectors across multiple machines.  The
> performance collecting daemons store their data locally (but we still
> allow access to the data through one central interface).
> 
> The event processing daemons (syslog, snmptrap, win eventlog) can also
> be distributed but they send events to a centralized relational
> database.  This setup can process large numbers of events.
> 
> Finally the core configuration database is also central but modeler
> daemons can be distributed.
>>
>>> 2) Can Zenoss aggregate views from several monitoring servers into an
>>> overall picture (similar to Nagios' distributed monitoring
>>> configuration)? For example, can a site-based monitoring server be
>>> configured to push it's results up to a "master" server somewhere?
>>>
> 
> The result of the things I describe above is a system that can scale to
> large numbers but has one integrated console to view, configure and
> manage your information.  I think this is a little different than a
> distributed nagios setup (although I'm not a nagios expert! :)
> 
> -EAD

I look forward to trying it. Thanks again for the great information.

- -=Tom Nail
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFFpmbw1zsLRjt/NQ0RAkK4AJ9CgZ8DbYY6dTllGmqEmUM75b2ZfQCfVR0i
Q8LJ8ZW6HMtzetPRRN5FYFo=
=35sh
-----END PGP SIGNATURE-----


_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to