I need some help troubleshooting frequent failures of OWA and Outlook. Exchange version is 2010 SP1, Outlook is also 2010 SP1. I believe from the research I was able to do that they are related to a failure of Exchange Active Directory Topology service.
For several days now, approximately every 12 hours or so the event log shows an error with AD Topology:
Process MSEXCHANGEADTOPOLOGY (PID=2364). When updating security for a remote procedure call (RPC) access for the Microsoft Exchange Active Directory Topology service, Exchange could not retrieve the security descriptor for Exchange server object SERVERNAME- Error code=80040934. The Microsoft Exchange Active Directory Topology service will continue starting with limited permissions.
This failure seems to cause a cascading failure of access to the server via OWA and Outlook.
Outlook displays a "need password" request and OWA displays the following message: "no server can be found with your account information."
Restarting the Microsoft Exchange AD Topology service fixes it temporarily.
One other piece of information that may be useful is that when logging on to the server in question with RDP, the Network icon at one point showed "no Internet access." However, the gateway and DNS are properly configured and I can access the Internet just fine. After rebooting, that message went away. Additionally, the problem did not occur at all over the weekend, so it does seem to be related somehow to relatively higher usage of the server--however, a few days in a row they happened around 7-8 PM and 7-8 AM--not high usage times in our office. Today it happened around 1:00 in the afternoon.
I've tested a few basic things already, but I'm stuck on how to resolve this. The CAS server is a VM on Hyper-V. I've tried switching the physical NIC in use with no change. I had much earlier applied the relevant hot fixes for network issues under heavy load on the Hyper-v server.
There were a few changes on the network recently. The same day that the problem first happened, I had removed the hub transport role from a server that I had been trying to do some tests on. Earlier in the week, I upgraded our domain functional level to 2008R2 in order to activate the AD Recycle Bin.
We have a network monitor in place--Zenoss--and the machine and all of our domain controllers haven't experienced any loss of connectivity at least as far as ICMP anywhere near the time of the failures.
This is a huge issue for our users. I hope that someone else out there can help me with this problem. Thanks!
Edit: just wanted to add I did find this link, which was the first in Google: http://technet.microsoft.com/en-us/library/dd789571(v=exchg.80).aspx however, there's only one VM NIC associated with the CAS and it is properly the first one in the binding order.