Redirect Outlook From Direct Server Connection To CAS Array

Microsoft has been always advice customer to configure CAS Array although there is only one server needed, unfortunately not many people is following this best practice.

My customer is going to deploy HA for their environment, which originally consist of 1 CAS/HUB server and 1 Mailbox server. Now they planning to have 1 more CAS/HUB server and mailbox server for HA purpose.

All the client is currently connected directly to the CAS and when  the new CAS is in place, I need to redirect them to connect to CAS Array.

There is around 700 users in the organization and I’m planning to perform the migration in a big bang strategy and minimize user’s downtime, so, I achieve the goal by using below steps:

  

1. Delete the old CAS DNS A record.

2. Create DNS A record for CAS Array.

3. Dismount all databases and re-mount them.

4. Outlook will perform a new autodiscover search and connect to CAS Array.

I hope this could help you to perform the migration with minimal user impact if you have a similar situation like this.

Deploying Lync Mobility With TMG Single NIC Configuration

I finally managed to deploy Lync mobility service today for my company's Lync server ever since Lync mobility was introduced few months back. You can find many useful deployment guide over the internet if you do a search, but what I would like to share is some experience about deploying Lync mobility services with TMG single NIC configuration which I found it is quite rare since most of the guide you found is mostly about having 2 NICs TMG configuration.

In my environment the TMG server is already in use to publish Exchange web services(OWA, Active Sync, Outlook Anywhere) . Since this is a single NIC TMG, you can only create one listener. If you try to create another listener for Lync services,  an error “A web listener specifying the same port and similar IP Addresses already used by the rule “[Your_Exchange_Services_Rule]”. The port and IP addresses specified in a Web Listener cannot overlap with the IP addresses specified web listener already used in a different rule”

To overcome this issue, simply use the same listener with a SAN certificate that contains both Exchange and Lync services FQDN.

For example, my original SAN certificate for Exchange is inclusive of webmail.domain.com , autodiscover.domain.com. I regenerated a new certificate in Exchange server with additional SAN which is required by Lync mobility services - lyncdiscoverinternal.domain.com , lyncdiscover.domain.com and etc.

After that, I went ahead to request new certificate from Lync Front End server with the same SAN (inclusive of Exchange web services FQDN) and we are almost good to go.

In TMG server I removed the original listener, and created a new listener by using the certificate which is newly created (In my case I export the certificate from Exchange server), followed by creating a new publishing rule for Lync discovery. When the rule is created and the moment I clicked "Test Rule" , it is showing a positive passed result and Lync mobility is ready to go live!

This is just an overall concept on how to deploy Lync mobility with TMG single NIC configuration, I hope this will at least provide you with some ideas if you have the same environment as mine.