Hello vCenter Community,
I have a vSphere environment with two hosts, each with 64GB of ram and dual Xeon E5-2620 CPUs. The environment was recently set up (last 5 months) and has been running with no problems so far. I am currently testing out the Horizon Suite, specifically Mirage. I have 4 Mirage VMs (management server, management console, mirage server, and database server). I also have a VM on each host for Starwind as well as the vCenter server, 3 test desktops, and a resource hungry web/DB server. My Hosts are running ESXi 5.5.
Host 1
Horizon Management Server - Server 2008 R2
Horizon Management Console - Server 2008 R2
Horizon Server - Server 2008 R2
Horizon Database Server - Server 2008 R2
Virtual Desktop 1 - Win7
Virtual Desktop 2 - WinXP
Virtual Desktop 3 - WinXP
Starwind Server - Server 2008 R2
vCenter Server - Server 2008 R2
Host 2
Starwind Server - Server 2008 R2
Web/DB Server - Server 2012 R2
Now, my problem...
On April 21st, I was working on testing my Horizon Mirage setup, which I had been doing for the past 9 days. I left work at 4:00pm. On April 22, I came into work at 8am, and logged into my vCenter server. When I opened a remote console to my Horizon Mirage Management Console, I noticed that Mirage wasn't running any more; I started it up and found that it was not configured at all... I looked at the snapshot manager for the Management Console and found that my Server had been reverted to the first snapshot I made (2 snapshots ago). I checked out my other VMs and they were all reverted to prior snapshots as well on both hosts. I checked the tasks/events log in vCenter and found no events between 11am on April 21st, and 8:40am on April 22. The last event shown in the Host 1 event log appears to be a DRS automated event, where DRS powered on my Mirage Management Server at 8:37 am on April 22. I checked the performance records for the last day and found that there was a performance spike on all of my VMs at 3am on April 22, pushing both servers to nearly 90% CPU usage and 80% Memory usage. The only thing I can think of is that Microsoft pushed out an update which caused the performance spike, but even that shouldn't have maxed the resources on all my VMs. I am still relatively new to vSphere management, so if I left out any pertinent information, please let me know. I will provide any logs necessary upon request.
Thanks.