I have been told that ESX can not accurately monitor its own Physical CPUs or Physical Memory conditions...Ie. if there is a fault in the hardware. Is that true? Thought I might be able to get some idea of the conditions of the physical memory and CPUs via tools like ESXTop
It depends what you mean by monitor hardware conditions - Do you mean the usage levels or CPU/RAM or some kind of proactive self-test?
If it's just the usage levels then esxtop and top will help. If it's a kind of self test then I'd be looking at manufacturer agents like HP Insight Manager agents or the equivilent to send SNMP alerts.
For esxtop check out this document http://communities.vmware.com/docs/DOC-9279 and for top just use a man top command
Thanks,
Neil
what I was referring to was actual hardware failures...like if one of the memory modules failed...how would ESX inform me? Or, if one of the Quad CPUs fail..... From what you are saying, it sounds like ESX really can't tell me. I'd have to rely on a third party utility
what type of hardware do you have? If running ESX classic most hardware vendors make available for free, agents. We have HP, so we use Insight Manager. But Dell offers thier tools (not sure of the name) as well as IBM, which I believe they use IBM Director
....aslo you can use the Health Status, if using 3.5, check under the configuration tab.
Troy
We have Dell and OpenManage. Was just thinking that ESX may have its own monitoring utilities. No worries. We have been using OpenManage,too
ESX has a CIM provider included that can tie into hw management utilities. So, the functionality to monitor is there, but is typically not as robust as individual vendor software packages. As pointed out, you can go into the health status view under configuration to see what ESX can provide for you out of the box, but you can't base alerts other than health status green/yellow/red as you can with the other 3rd party packages.
-KjB
VMware vExpert
Don't forget to leave points for helpful/correct posts.
I'm a little late to this party, but for anyone else reading this down the road...
If you google ESX and SNMP there are plenty of guides out there for sending SNMP traps to a syslog server that can better present the data. I did this for all of our ESX servers and it works great. What I have below is what worked for me:
http://www.tooms.dk/?page=http%3A//www.tooms.dk/forum/topic.asp%3FTOPIC_ID%3D127
The only difference was my last steps after I configured the file were this:
1. Once the file is updated to match, restart the snmpd service
service snmpd restart
2. Ensure that snmpd service is allowed through the ESX firewall by running
esxcfg-firewall -e snmpd
3. Logout of the server
SNMP is one way, but CIM is another good source of information and this info is exposed in the APIs, especially important when talking about ESXi.
Here is a script that'll allow you to monitor all hardware components and also provide emailing capabilities, it works on ESX(i) 3.5 and 4.0:
=========================================================================
William Lam
VMware vExpert 2009
VMware ESX/ESXi scripts and resources at:
VMware Code Central - Scripts/Sample code for Developers and Administrators
If you find this information useful, please award points for "correct" or "helpful".