exchange 2010 database cross-site failover

Hi,
I'm working on Exchange 2010 and created a lab with the following setup:
AD site -site1 subnet 10.91.50.x/24
ad/dns server (1. witness srv)
test1 win7 system with outlook 2010
test-exch1 ht/cas/mb with (db1 and db2 mailboxes)
dagname dag1 -db1 is active, db2 is copy on exch1

AD site -site2 subnet 10.93.50.x/24
ad/dns server (alternative witness srv)
test-exch2 ht/cas/mb with (db1 and db2 mailboxes)
dagname dag1 -db2 is active, db1 is copy on exch2
test2 win7 system with outlook 2010

Details: using single namespace (webmail.test.ca) as an external url for both system and server names are also in the cert. (these are all in test/lab environment)

when I powered off test-exch2, DAG switched the "db2" database over to test-exch1. I've also changed the rpcclientaccesserver to test-exch1 for that database. However, test2 system which is in site2 is still trying to connect test-exch2 server (failed one). It seems still querying AD and getting test-exch2 scp record from the ad-site2 server.

Is there any way to automatically update scp records for the remote site in case of database failover? Or maybe, better to go different namespace for each site?

thanks,
Cem
c-iscenAsked:
Who is Participating?
 
Simon Butler (Sembee)Connect With a Mentor ConsultantCommented:
CAS Array takes about two minutes to setup.
The DNS change takes 10 second to do in the event of a failure.

That is how you are supposed to manage site failover for clients. No other method.

Simon.
0
 
Simon Butler (Sembee)ConsultantCommented:
The DAG only protects the mailbox database, it does not protect the client.

For failover, you should configure an RPC CAS Array, which is AD site specific. Then when you failover, reconfigure the RPC CAS Array address to point to the CAS role holder in the second site. Cut the TTL time down to 10 minutes so that the change takes effect quickly.

Changing the RPCClientAccessServer value is not the way to change the server the client connects to - that is what the RPC CAS Array feature is for.

The RPC CAS array must be a host name that does nto resolve externally and should not be in the SSL certificate. It should be a host name unique to that feature. The most frequent name for the CAS array is outlook.example.local.

Simon.
0
 
c-iscenAuthor Commented:
Simon,

I'll give a try but in my case I have only one CAS server on both AD sites. So, instead of creating CAS Array would not be easier to change test-exch2(failed server) dns to point other site's exch server?

Cem
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.