Hi
Hope someone can help. I am working on an Exchange 2003 to 2010 migration, which will then quickly move onto a 2010 to 2013 migration and need some clarification on the namespaces to use. I am aware that if I do not do this right at the 2003 to 2010 migration, this will cause a headache at the 2010 to 2013 migration.
Some background:
- 2003 Functional Level Domain - 2 x 2008 DC's
- Currently users are on a 2003 exchange cluster with a mix of RPC (internal users) and RPC over HTTP connections (roaming users)
- We will be installing Exchange 2010 on a single server, with CAS, HUB and Mailbox roles and no load balancer, as we will be moving quickly to 2013.
- We have two Kemp load balancers ready for Exchange 2013.
- Exchange 2010 is installed on a single server (exh2010.domain.local) and configured with an CAS array name (exh-cas.domain.local) which is resolvable internally only.
- Currently we have multiple smtp namespaces e.g. @company.com, @company2.com.
- Our main website etc is www.company.com
- Our public facing services are at https://service.mycompany.com
- Our 2003 RPC address is https://webmail.mycompany.com
I understand that the 2010 RPC CAS array name should be separated from the Outlook Anywhere (RPC over HTTPS) address so that when 2013 takes over the HTTPS address, the RPC connections are not broken.
Two Questions:
- Do we have to use the HTTPS same namespace for 2013 as we do in 2010? Its just I would want to test the Kemp load balancers before making them live (slow careful transition), and giving them a different namespace, e.g.https://mail.mycompany.com would allow a migration, rather then a cutover.
- Can we use the *.mycompany.com address rather then the company.com address, even though we have no SMTP addresses at mycompany.com? Can autodiscover still work?
Thanks in advance for any guidance
Cheers
Steve