Thanks for the reponses, nice to see we have an active community here. There
were enough hints here for me figure out I have a BIG disk I/O issue.

The box I am using is also collecting several gig of syslog a day, however
this doesn't seem to be a problem since the cpu and memory load is LOW when
vmware is not running.

It's only when vmware is fired up that problems begin. When zenoss is
started cpu jumps to 100%. Most of that is shown by 'top' to be system cpu
usage. The interrupt load goes above 2.0. My understanding is that these are
good signs of load on the hard drive.

I wanted to isolate what part of zenoss was causing the problem so first I
followed Eric's suggestion and added 'RouteMap' to the ignored plugins. No
change, but suspecting it was still to blame I started with a fresh virtual
machine and modified the routers class before adding my devices.

This seemed to fix the immediate problem. Now when zenoss is started the
first couple of minutes are 100% cpu but then it settles down. Also when
devices are added cpu naturally jumps but then it settles down again. I
conclude the RouteMap plugin needs some work! Eric are there any details
available on what this plugin is doing in the backgroun?

However I still find the system load quite excessive and would like to find
out if this is just a feature of vmware itself or if I need to source
another hard drive for when I install zenoss natively.

What to other people see in top when discovering devices on a similar
machine?
(Dell PowerEdge 2650, 2x 2.4GHz xeons, 1Gb Ram, 1x160Gb HD)

cheers,
Ben



On 3/27/07, kerneld <[EMAIL PROTECTED] > wrote:

This article is nagios specific i.e there could be issues with parallelism
on nagios, or on that specific platform.
Zenoss doesn't seem to have any problems with having 2 virtual processors,
and will run alot slower with a single CPU.

One thing I forgot to mention is that on my VM instance, I have a bond
interface that is configured using 802.3ad . the interface spans 4 gbit
NICs and the corresponding ports on the switch are configured in an
etherchannel group. this lowers CPU utilization and reduces IO wait. you may
want to consider similar things.

------------------------
Mohamed Hussein




-------------------- m2f --------------------

Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=4897#4897

-------------------- m2f --------------------



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




--
---------------------------------------------------------------------------------------------------------------------------------------------------------

"Because it that the times revive as time is fresh somehow, and it to feel
wins why, and, as for it, all forget an old thing" - Japanese saying
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to