2013-11-12 03:57 PM
As others have also noted, I am seeing the VSC service account log into vCenter every 15 seconds. This is filling up our logs as well. Does anybody know if this is expected behavior?
We've tried re-registering VSC wtih vCenter and that didnt fix the issue as it did for somebody else. Another thread indicates that this is normal behavior however I spoke to our NetApp rep and he said he doesnt see this in his customer environments.
Solved! SEE THE SOLUTION
2013-11-12 05:42 PM
That is currently expected behavior. Optimization and Migration polls every 15 seconds for virtual machine events. Other than the excessive logging, is this causing an issue in your environment?
2013-11-12 06:05 PM
It is not an issue in our environment however it certainly seems excessive. If it is expected behavior, it is strange that our NetApp representative didn't see the same behavior in another one of his customer environments who he claimed to be a heavy VSC user.
2013-11-12 06:12 PM
The 15s polling is related to a feature added in VSC 4.2, so it's likely that customers running VSC 4.1 or earlier would not necessarily see that behavior. We can look at possibly making this configurable going forward, or feel free to open a support case.
2013-11-12 06:21 PM
Thanks for the response. And you're certain it would manifest itself as a login/logout to vCenter? Here is what I see in the vSphere client logs when looking at the Events at the root:
User <Domain>\<VSCServiceAccount>@<vscServerIPAddress> logged in
User <Domain>\<VSCServiceAccount> logged out
I did open a ticket as well. If anything, hopefully it will be tracked as another feature request to make this configurable.
2013-11-12 06:30 PM
Yes, that periodic job currently logs in as an admin, retrieves the VM events since the last time it checked, processes them, and logs out. Possibly, this can be changed in the future to remain logged in for a longer period of time which would reduce the amount of logging, or the job interval can be configured to be longer.
2014-04-17 09:21 AM
I opened a ticket with NetApp and maness is correct - the 15s polling interval is by design.
"The 15s polling is related to a feature added in VSC 4.2, so it's likely that customers running VSC 4.1 or earlier would not necessarily see that behavior. We can look at possibly making this configurable going forward, or feel free to open a support case."
2014-05-06 09:21 AM
Maness, that would definitely be a welcome improvement.
I've just tracked a Virtual Center crash issue down to the VC database being filled by too many events (SQL Server Express database, with a 10GB limit and 8.5GB used for events tables).
Half of them came from VSC.