A physical move of your entire exchange environment will require changing gateways, subnets, and IP addresses for everything. Here’s how to change the IPs of everything involved if you are using Exchange 2003 in a cluster environment.
To get started, here are the pieces to the puzzle. You may have more cluster servers:
- CLUST01: Windows Server 2003 – Cluster Server (Node) 1
- CLUST02: Windows Server 2003 – Cluster Server (Node) 2
- CLUSTER: The IP Address of the Cluster Server.
- EXCHANGE: The Exchange server that bounces back and forth from cluster to cluster.
These are the names of the IP Addresses you’ll need to change:
- Node 1
- Node 2
- Cluster IP Address
- Exchange IP Address
And here’s the steps we’ll be taking:
- Move resource groups to Node 1
- Take all resouces groups offline
- Put Cluster Services in Node1 and Node2 into manual
- Power everything off and move it
- Power Node 1 and Node2 on and change NIC IP Address information
- Change Cluster IP Address
- Change Exchange IP Address
- Update Queue Information in Exchange
- Update DNS on your Domain Controller(s)
Move Resource Groups to Node 1
You cannot begin without doing this. You’ll be in a world of hurt if you don’t know where your resource groups are once you start making IP changes. You don’t have to do this step if the resource groups are all on Node1 already. Here’s how you do this:
- Log into Node1 with admin rights
- Open up Cluster Administrator
- Head to Cluster > Groups > Cluster Groups and right click
- Choose “Move Group”
Here’s a screenshot to show right where that is.
Take all resource groups offline
While in the Cluster Administrator:
- Head to Cluster > Groups > Right Click on Cluster Group and click Take Offline. If things go well, all of your resource groups will become offline.
Put all services in manual
I like to do this now so I don’t forget later on. We don’t want these services starting up at the new location without us making more changes once we get there. Navigate to services and change the status of the Cluster Services to manual so it doesn’t start up upon arrival.
Power everything off and move it
One small tip, be sure and know exactly how the pieces of this puzzle get put back together physically. Odds are you have multiple NICs (Which I like to label on the outside) and other things you’re not starting at on a daily basis. Document everything!
Power Node 1 and Node2 on and change NIC IP Address information
Odds are you have a separate network for communication (heartbeat) between nodes. For my environment, this didn’t have to be changed as there was no overlap or issue. But the production NICs are what will need to be changed to communicate with the new gateway on the new subnet. This is as easy as changing normal IP Addresses, and because our Cluster Service isn’t started, we don’t have a reason to worry about breaking anything.
Change Cluster IP Address
The Cluster IP Address is another IP that is really straight forward to change, but it’s located under the resource group called Cluster IP Address:
- In Cluster Administrator, Head to Cluster > Groups > Cluster Groups > Double Click “Cluster IP Address” in the list of resources to the right > Parameters.
- Change the IP Address to the new IP you have designated
- Apply and close
Here’s a screenshot of what that parameter looks like. Nothing scary here.
Change the Exchange IP Address
This is the big one, only because other place will need this IP Address, like in your queues. This is changed in the same way the Cluster IP Address was changed. Look in the list for Exchange IP Address and change it under the parameters tab.
Update protocol information in Exchange
- Open up System Manager for Exchange
- Head to Administrative Groups > First Administrative Groups > Servers > Your server you just moved > Protocols
- From here, head to all of the protocols that your server you just moved is responsible for. For us, this server was only responsible for HTTP and SMTP. Right click on each of them and change the IP Addresses there.
- There is an advanced button for some of these protocols, you’ll want to make sure the IP Address is changed there as well.
A screenshot is above of the protocols. Again, make sure advanced is changed as well.
Update DNS on your Domain Controller(s)
You could have done this earlier if you wanted to, but make sure it gets done or nothing will work.
Bring everything online in the Cluster Administrator, make sure everything can come online without errors. Also, make sure your queue for that server is flowing without issues or getting backed up. When everything is online and Node1 looks good, move the group to the other cluster and make sure things work over there.
Let me know how it goes and if there’s anything I’ve missed.