Hi there,
We have a setup as below..
Flint, US
3 x CAS servers in an Array
Dublin, EU
3 x CAS servers in an Array
Each server has got outlook anywhere enabled identically with a front end name and SSL certificate of outlook.domain.com. This DNS record points at the US CAS array. What has happened for the last 12 months is all users connecting via outlook anywhere connect to the US, then exchange itself will route the request to the correct mailbox regardless of it being in the US or Dublin.
This past weekend we deployed SP2 RU6 onto all of the servers. Since the weekend, any of our European users can now no longer access via outlook anywhere. We get the Server unavailable error message. I have done lots of research and came up with the same proposed solutions. Which include disable/re-install rpc proxy component/enable etc. None of this has worked.
All access for US users is fine. All other access for EU users such as OWA, EWS, ECP is fine. If I check via testexchangeconnectivity.com with the credentials of a EU user, ALL tests pass succesfully so appears to be no issue.
Now I think I have narrowed it down a little. If on my machine I modify my hosts file so that I point at each of the Dublin backend servers individually then they all work perfectly. If I then modify my hosts file to point at each individual back end US server, no joy. So to me it looks like whatever has happened after installing RU6, exchange is no longer now taking my connection in Flint and pointing it correcly at my mailbox in Dublin.
I am really struggling to see what else to check/change here now. Can somebody throw any light onto this? Could this be a bug in the RU6? If possible, I would like to see if we can resolve first before having to roll back.
Thanks, John