0

I have an Exchange 2010 SP3 DAG with 7 servers (3 in production, 3 in DR and a LAG). I have 2 CAS Arrays simply named casarray1 and casarray2. casarray1 sits in production and has the three production servers on it whereas casarray2 sits in our DR site and has the three DR servers connected to it. All servers sit behind 2 Kemp load balancers in each location.

Part of our DR test is to create a temporary database in DR and replicate it to another DR server, which I've done and named 2014DRTest. I picked 15 mailboxes from our IT department and scripted their move request, which went fine however once their moves are completed, the Outlook client won't connect to Exchange. I have multiple versions of Outlook testing, 2013, 2010 and a 2007.

I moved everyone back except myself for testing purposes.

I ran Get-MailboxDatabase | fl Identity, RpcClientAccessServer and I show that the 2014DRTest database is pointing to the correct CAS Array.

Looking at the properties of the Outlook account, the server alias now shows the DR CAS Array, but launching Outlook simply stays on "Trying to Connect"

This same test worked perfectly with no issues last year and the only thing that's changed was rolling up from SP2 to SP3.

I'm unsure which rabbit hole to go down to look for the root cause. I need, for audit purposes, to show that a failover would work and with this behavior, I don't believe that a failover would operate as intended should we lose a production server.

Dave H.
  • 125
  • 1
  • 2
  • 10

1 Answers1

0

We uncovered an incorrectly set Default Gateway on one of the DR servers' MAPI NIC Team.

Michael Hampton
  • 244,070
  • 43
  • 506
  • 972
Dave H.
  • 125
  • 1
  • 2
  • 10
  • Feel free to mark this as an answer when the time limit expires, as this could be a useful snippet for others who are having a similar problem. Default gateway/routing problems are common enough in DR situations. – Mark Henderson Oct 16 '14 at 23:07