Solved

Exchange 2010 2 Server DAG - Do I really need CAS Array?

Posted on 2011-03-02
11
1,166 Views
Last Modified: 2013-05-01
Ok, I have 2 Exchange 2010 severs clustered as a DAG.  Explain this to me like I'm a 6 year old: with 2010 clients and outlook anywhere web based clients why would I need a CAS Array? Newer Exchange clients attach to the RPC service and not directly to the CAS.  If one CAS goes down they automatically find the next RPC server.. correct?  If not why wouldn't the DAG internal IP address assigned to an MX record work?

Also, if you assign a CAS array I've read you don't need NLB...?
0
Comment
Question by:edutechceo
  • 5
  • 4
11 Comments
 
LVL 49

Accepted Solution

by:
Akhater earned 500 total points
ID: 35024285
1st in exchange 2010 RPC connections are terminated directly on the CAS servers

>> If one CAS goes down they automatically find the next RPC server.. correct?
No not correct at all, in outlook configuration you have the name of your exchange server configured, if this is pointing to CAS server1 and this one goes down outlook will not look for any other

>>Also, if you assign a CAS array I've read you don't need NLB...?
wrong

Let me try to explain things in my own terms

a. Each Database has a property RpcClientAccessServer this entry points to the name of the CAS server responsible for this database. So when outlook is being configured autodiscovery service will fill outlook with the value in RpcClientAccessServer of  the database where the user's mailbox resides

b. technically speaking a CAS Array is nothing but a unified name you give to all CAS servers in a specific site and then change the RpcClientAccessServer property on all databases in this site to point to the name you have just created. the purpose, as you might imagine, is to make outlook connect to a unified name no matter how many cas servers you have in this site

c. Now that you have your casarray name you need to make this name resolvable by dns to the CAS servers in this site and here you have 2 options
    1. NLB: you setup an NLB between all these cas servers and you make the fqdn of your cas array point to the virtual ip of the NLB. however you should know that you cannot use Windows NLB if your CAS server is on a DAG member. if this your case you need to use hardware Load Balancer
    2. DNS Round robin: if you cannot use NLB you can use dns round-robin ( a lesser solution) which means  having 2 dns records in your dns for your cas array name each pointing to the IP of  one of your cas servers

Hope it makes more sense now
0
 

Author Comment

by:edutechceo
ID: 35026424
What about pointing the CAS array to the internal clustered IP address of the DAG?
0
 
LVL 49

Assisted Solution

by:Akhater
Akhater earned 500 total points
ID: 35026894
you got the idea right but it won't work or at least not fully

it will give you high availability for your CAS but not load balancing
0
 

Author Comment

by:edutechceo
ID: 35027671
Ok - so it will try and connect to the DAG server that is up if one fails (CAS) if connecting to the DAG internal IP cluster network then?  If so this appears viable even without NLB - after all with 2 servers at the same site and 50 clients I'm not sure how NLB would be any better.

0
 
LVL 49

Expert Comment

by:Akhater
ID: 35027748
it means that when the 2 are up only one will be serving the 50 clients and not 25 clients on each as it would be the case in NLB
0
IT, Stop Being Called Into Every Meeting

Highfive is so simple that setting up every meeting room takes just minutes and every employee will be able to start or join a call from any room with ease. Never be called into a meeting just to get it started again. This is how video conferencing should work!

 

Author Comment

by:edutechceo
ID: 35028555
Makes sense - obviously spreading the load out with 50 clients is not really necessary - so if the clients will automatically find the other CAS in the DAG internal network (from the CAS Array assigned to the DAG internal net) then great.  Unless I'm missing a major downside to this.  At this point I guess round robin isn't necessary since it's one IP address...?
0
 

Author Comment

by:edutechceo
ID: 35029423
So is this correct - Round Robin isn't necessary since I'm pointing to the DAG internal cluster network IP or am I fooling myself here?  
0
 
LVL 49

Assisted Solution

by:Akhater
Akhater earned 500 total points
ID: 35031930
in that case no need for round robin no just point the cas array name to the dag ip
0
 

Author Comment

by:edutechceo
ID: 35032109
That's what I'll do!  I'm going to implement this weekend...
0
 
LVL 74

Expert Comment

by:Glen Knight
ID: 35356744
This question has been classified as abandoned and is being closed as part of the Cleanup Program. See my comment at the end of the question for more details.
0

Featured Post

What Is Threat Intelligence?

Threat intelligence is often discussed, but rarely understood. Starting with a precise definition, along with clear business goals, is essential.

Join & Write a Comment

Utilizing an array to gracefully append to a list of EmailAddresses
Find out how to use Active Directory data for email signature management in Microsoft Exchange and Office 365.
In this video we show how to create a Resource Mailbox in Exchange 2013. We show this process by using the Exchange Admin Center. Log into Exchange Admin Center.: Navigate to the Recipients >> Resources tab.: "Recipients" is our default selection …
The video tutorial explains the basics of the Exchange server Database Availability groups. The components of this video include: 1. Automatic Failover 2. Failover Clustering 3. Active Manager

759 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

20 Experts available now in Live!

Get 1:1 Help Now