In this blog post we will talk about OpsMgr 2012: Hyper-V Management Pack Extension for Windows Server 2012 Hyper-V role. This management back came out on 22, June 2013 and was published on “Kevin Holman’s System Center Blog”.
This management pack extension were using the original Windows Server 2012 Hyper-V management pack with some extended capabilities to monitor the following properties of Windows Server 2012 Hyper-V hosts:
- VMs Integration Services Version monitor
- Hyper-V Replica Health Monitoring Dashboard and States
- SMB Shares I/O latency monitor
- Hyper-V Hypervisor Logical processor monitoring
- Hyper-V Hypervisor Virtual processor monitoring
- Hyper-V Dynamic Memory monitoring
- Hyper-V Virtual Networks monitoring
- NUMA remote pages monitoring
- SLAT enabled processor detection
- Hyper-V VHDs monitoring
- Physical and Logical Disk monitoring
- Host Available Memory monitoring
- Stopped and Failed VMs monitoring
- Failed Live Migrations monitoring
Problem:
Although everything look fine in the start but over the time, after first deploying this I have seen that occasionally “MonitoringHost.exe” and some other core Hyper-V related processes on my Windows Server 2012 Hyper-V hosts takes 100% of CPU time sharing. This situation stays same until and unless I manually go and kill the process or as a result of this if it stays for longer period we use to see that Hyper-V hosts get unexpected reboot.
Since the similar issue was reported in the RTM release of System Center Operations Manager 2012 agent, so I didn’t think about this newly installed community driven Hyper-V 2012 management pack extensions. Microsoft also released System Center 2012 SCOM roll up 2 in which they documented and fixed this problem. And the similar also got fixed in the System Center 2012 R2 release.
Recently one of my customer reported me that they waited for System Center 2012 R2 and when the initially upgraded their SCOM 2012 server with R2 and agents on their Windows Server 2012 Hyper-V hosts. Things didn’t get change and they remain seeing “MonitoringHost.exe” alike process to stays at 100% CPU utilization on their servers.
Cause and Solution:
After some troubleshooting we found that this is Hyper-V management pack extensions which is requesting a lot of data which agent sometime fails to provide or get stuck in between which causes its process to go beyond and stay at 100% of CPU utilization. Even if you install SCOM 2012 Rollup 2 or upgrade to SCOM 2012 R2 the issue will not get fixed.
So the solution is to remove Hyper-V management pack extensions from the environment and use the default as it is not created and tested by Microsoft but by some community driven people. If you don’t like the solution then you have to dig deeper and need to find the actual performance monitor which is causing this 100% CPU utilization and might get it disable in SCOM.
Cheers!
Azure Hybrid & Cloud Migration Blog 