Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have a file and print server running on Windows 2008 R2 fully patched in a VSphere environment (ESXi 5.1 fully updated). Every evening between 19:20 and 19:30 our monitoring software reported that the available memory is 1% and performance is dire. There is nothing in the event logs to point to an issue.

At this point in the evening I am general the only user on the system to check to see why these alerts are going off.

Things I have done;

  • Checked to see if any backups are running – None at all.
  • Checked Scheduled tasks – None before or during this time period.
  • Moved the VM to another host.

AV is disabled to rule out that as the issue.

The server does not have any problems during the day with memory when fully loaded with about 50 users.

The server did have 4GB ram provisioned but I have increased this to 5Gb. Running PrefMon at the time (I will save the graphs tonight) There is very little CPU usage at the time but RAM usage goes up.

EDIT : After lots of digging I have found that this is due to Shadow Copied being deleted. By the VSS Service. My problem is now that I cannot disable this. I have been through the VSS Admin documentation but cannot see how to change the time that the maintenance runs. Currently it is running at 19:15 GMT.

share|improve this question
What do the VSPhere logs show during this time, rather than the guest logs? – Dan Nov 26 '12 at 10:36
ah! Just looked at around the same time our Exchange Server is doing a snapshot for vDR. Ill move that and report back tomorrow. – t1nt1n Nov 26 '12 at 10:56
@Dan Just looked at the VMKwarning.log for that host and there isnt anything obvious. Apart from cpu11:8203)WARNING: ScsiDeviceIO: 1211: Device naa.6001405175a4a72da36bd414dd9ba4da performance has deteriorated. I/O latency increased from average value of 8879 microseconds to 729397 microseconds quite a few times during this time. Maybe the swap file has to much latency. – t1nt1n Nov 27 '12 at 9:48

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.