Right now I am just looking at overall host usage of the ESX servers, I still have client side opsview agents that follow the vm's. I just queried two of our older ESX servers that are about to be replaced and it took about 3 seconds on each one.
Here is an example check nag...@nms-cae-s01:/usr/local/nagios/libexec$ time ./check_esx3.pl -Hjekyll -l CPU -uvmsmonitor -p'xxxxxx' CHECK_ESX3.PL OK - cpu usage=3605.00 MHz(56.48%) | cpu_usagemhz=3605.00Mhz;; cpu_usage=56.48%;; real 0m3.034s user 0m1.330s sys 0m0.620s Can you share the sort of check you are doing and we can compare times? This particular slave server is checking 5 ESX servers and I haven't seen any problems with time outs yet. James Whittington VC3, Inc. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Jason Lavoie Sent: Tuesday, May 19, 2009 4:45 PM To: Opsview Users Subject: Re: [opsview-users] check_esx3 On 05/08, Jason Lavoie wrote: > On 05/08, James Whittington wrote: > > I have had success using check_esx3 by Op5 developer.. > > http://www.monitoringexchange.org/cgi-bin/page.cgi?g=2848.html;d=1 > > > > It requires that the Infrastructure Perl Toolkit is installed on your > > monitoring server but the install is very easy. I've packaged up the viperltoolkit for debian, and have this working now on our slaves. Unfortunately, it is amazingly slow -- a single "VM" check is 30+ seconds. It gets much worse as multiples checks are done. I have 8 checks running right now (as a test) and I still occasionally hit the (default 60s) nagios service_check_timeout. How is this plugin scaling for you? Maybe our vcenter server is undersized/overloaded... -j -- Jason Lavoie Ratvarre sbe uver [email protected] _______________________________________________ Opsview-users mailing list [email protected] http://lists.opsview.org/listinfo/opsview-users _______________________________________________ Opsview-users mailing list [email protected] http://lists.opsview.org/listinfo/opsview-users
