I'm migrating from Exchange 2010 to 2013 and have run into an issue.      My environment has ambiguous URLs (my internal CAS Array name and the Outlook Anywhere name are the same).  The recommendation from Microsoft (Brian Day) is to force all clients into HTTPS mode before cutting over DNS to Exchange 2013.     But, I also have a very heavy Citrix/Terminal Services environment which is recommends the Outlook client to be in online mode.

While testing the HTTPS mode with my pilot users, feedback has been very negative.   Intermittently, the client will be "slow".   If I look in Outlook Connection Status, I will see "Ave Resp" in the 150-200ms range.  If I simply turn off HTTPS and return to MAPI, the connection status is under 15-20ms and our users are happy. 

I've looked at performance on the Exchange 2010 CAS role and I'm not seeing anything that looks out of the ordinary?   Does anyone had success with placing Outlook 2010 on Exchange 2010 in HTTPS while in online mode?

You really need to fix the namespace issue.  Then move all your client access to go through Exchange 2013.  Once that's working satisfactorily, moving mailboxes shouldn't cause too many issues.

You don't believe these online mode RPC over HTTPS problems will exist on Exchange 2013?  We can't move to MAPI over HTTPS because we also have legacy public folders ...

• Edited by Thomacy 19 hours 3 minutes ago

You don't believe these online mode RPC over HTTPS problems will exist on Exchange 2013?  We can't move to MAPI over HTTPS because we also have legacy public folders ...

• Edited by Thomacy Friday, February 27, 2015 4:27 PM

Clients will connect directly to the mailbox server for public folder access.

Hi Thomacy,

According to your description, I understand that Outlook connect to Exchange sever very slow with Outlook Anywhere, however works fine in MAPI.
If I misunderstand your concern, please do not hesitate to let me know.

I notice that you are migrating Exchange 2010 to Exchange 2013, Exchange 2013 servers with the Client Access and Mailbox server roles are responsible for message transport in a hybrid deployment. Also, CAS array isnt used in Exchange 2013, and by default Outlook client use Outlook Anywhere to connect to Exchange 2013.
Therefore, we recommend change MX record, all virtual directory URL and point to Exchange 2013 server, then move mailbox to Exchange 2013. As Ed mentioned, clients will auto redirect to the mailbox server for public folder access, you also can move public folder to Exchange 2013. More details about Use serial migration to migrate public folders to Exchange 2013 from previous versions, for your reference: https://technet.microsoft.com/en-us/library/jj150486(v=exchg.150).aspx

Additional, the steps of migrate Exchange server to 2013, please refer to:
http://blogs.technet.com/b/meamcs/archive/2013/07/25/part-2-step-by-step-exchange-2007-to-2013-migration.aspx
 
Best Regards,
Allen Wang

Oh, I have ambiguous URLs, so, the recommendation from Brian Day is to switch clients over to HTTPS before cutting over DNS (http://blogs.technet.com/b/exchange/archive/2013/05/23/ambiguous-urls-and-their-effect-on-exchange-2010-to-exchange-2013-migrations.aspx).