VMs hung over some time

[ChipandDale]

Have some VMs running in ESXi HA cluster. All of them have same virtual hardware, same OS. Sometimes one of them hungs. It's always a random VM. I see a blue screen (wallpaper, NOT BSOD) on a console of hypervisor and nothing else. No any load on CPU, memory or disk is present. I can establish connection over telnet for port 3389, but can't connect over RDP. No any records in eventlog. After reset VM works properly.

VMs:

XP SP3, all security updates installed.

2 cores on CPU

1 GB memory.

[dhanarajramesh]

can u check those affected VM's are from the same ESXi? I hope u have may received host not responding event in tasks and evet tab. I had faced this issue, after deep diagnosing I found that Physical memory leak issue and given memory controller issue in esxi level.

[ChipandDale]

Checked all of this. 2 machines are on one host and third is on another (i have 3 hosts at cluster). No any alerts at hosts logs or tasks. Last action written before the machines hang is  about successfully taken snapshot and backup. Backups are made using NetApp tools

[dhanarajramesh]

i hope this is what hapening on your VM's. refer this link. http://www.veeam.com/kb1681

[ChipandDale]

We don't use Veeam backup. The only backup are NetApp snapshots on storage. And there may be a heavy load, cause it runs in the night with some other apps on VMs. I'll check this.

[homerzzz]

What applications are running on the VMs? Sounds like a memory leak within the guest. Check non-paged pool usage for the processes over time.

We use Netapp snapshots and I have never had them hang the VMs.

[macvirtual]

Hi,

I sometimes face that kind of problem, where some VMs hang but still able to connect certain port.( nothing comes out from that)

Generally, these kind of phenomenon are cause of storage. When I see those kind of issue, I would check storage first, if it can connect properly, if able to readble/writable.

So, my advice for it is to check your NetApp storage.

Best,

MAC

[ChipandDale]

We have about 50 VMs. Some are on Unix-like OS, some are on windows. We have problems only with VMs running windows.

Some of VMs that hang run some server apps, some runs nothing except of OS (admin machines for apps on other VMs).

[ChipandDale]

Like i wrote above we have many VMs, they all use this only storage. Only one VM hungs at a time. Though storage was unavaliable we would have all VMs fault.