VEEAM Causing WMI Memory leaks

[Link]

Hi

I have had a case open now with VEEAM Support for over a month.

VEEAM monitoring is causing excessive WMI Handles which causing a memory leak on the WMI Service which causes it to stop and restart frequently.
This is preventing data collection and is also causing server hangs from time to time.

I have doubled the WMI Handle Count as well as made several adjustments to the registry to adjust the polling time to 3 minutes instead of 60 seconds, yet both of my clusters are still experiencing high WMI handle counts.

I have two Hyper V Clusters running on Server 2012. Both are experiencing the same problem when being monitored by VEEAM One.

My event log is flooded with events showing the WMI service restarting constantly, as well as Handle counts in excess of 90 000 (default value is 4096)
I have logged a call with Microsoft and they have confirmed WMI is crashing due to too many handles.

VEEAM suggests increasing the WMI handle count 12 fold, but MS has advised against this.

I am unable to use monitoring via Remote Registry as my clusters are located in different domains with selective trusts only.

Has anyone else experienced high WMI Handle Counts / Memory leak problems as a result of VEEAM One monitoring?
Any recommendations or feedback would be appreciated.

[Vitaliy S.]

Hello Leigh,

I haven't seen this issue reported before, but I wonder what happens if you increase the number of WMI handles and try to run any other application (or test) to see whether WMI service keeps crushing or not?

Thank you!

[Link]

I have doubled the WMI handle count, and this just delays the memory leak slightly.

Handles are set to 8192.. VEEAM is generating handles in the 95 000 - 165 000 range.