SIM and VMware Traps

[sterls27]

We currently have Insight Manager and VMware Virtual Console integrated where VC can forward traps to Insight and Insight sends out email alerts. The emails are received fine and I imported the MIBs from VMware into Insight Manager so it would recognize the traps. My question/issue is when the traps are sent out via email, such as a VM status has changed from green to yellow within VC, it does not include any information within the email. If you go into Insight Manager, it will show you the VM, what status changed, what the counters are, etc. The email only states:

Event description: This trap is sent when entity status changed.

Did I miss something and variables need to be added to the cfg file for the MIB? How do I get the same information within Insight Manager details to be displayed within the email? Any help would be appreciated.

[Aravindh Rajaram]

What's the format chosen by you while creating the automatic event handling task?

[sterls27]

The alerts fall into the same category as the System Event Handler, we do not have a custom task defined for these alerts. The format for the Sys Event is Standard, and other traps that use the Sys Event task are forwarded with information included (i.e. Critical Threshold warnings, iLO resets, etc.).

[Rob Buxton]

Not answering your question, but I'd be interested how you get on. I found that VC seemed to be quite sensitive to changes. e.g. it only takes the briefest of spikes to generate the error.
I must admit I've not tested this recently though.

[sterls27]

Actually, we do receive a lot of alerts, but you can set the thresholds within virtual center for memory and cpu same as you can for Insight Manager. So it's a matter of tweaking the alerts so you don't get innundated.

[Rob Buxton]

Yes, but it seems the briefest of spikes that crossed the threshold would trigger the alert.

In the HP Agents it seems they're a bit less sensitive to spikes, so a CPU spike will not trigger an alarm whereas a prolonged CPU burst will.

I just found that the number of alerts I got from VC made it less than useful here.

I might see if there's a way of modifying the VC behaviour - I might try the VMWare forum.

[sterls27]

I see your point, however we have the cpu alerts disabled to send traps and have the cpu thresholds set within Insight Manager as we have the VMM pack installed. We don't use the VC (Virtual Center) to monitor the CPU, mainly memory usage and the Hosts connections within VC. We were alerted by VC when one of our ESX hosts purple screened before Insight Manager noted it was down. Only a 2 minute difference, but it was still nice.

[Rob Buxton]

Yeah, we're using VMM as well. Have you seen the thread regarding VMM, it seems to cause problems with vmotion at some sites . Ole, who also posts here and in the vwmare forum, has posted a few items on it. He's had to disable it. At the moment I've still not re-integrated VMM with VC. I'm waiting to see if anything comes out about it.

[sterls27]

The VMM agent at least notifies us if the VM's are using extensive CPU/Memory. I've also set up SNMP Informant that does WONDERS for reading SNMP thresholds you can define. I've pretty much disabled the VC alerts as well and am waiting to see what wonderful things we'll be able to do with Virtual Center 2.5. Thanks for the suggestions!