• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 2503
  • Last Modified:

Exchange 2010 CAS Array and FailOver Clustering

I have a detected server that will only host my Exchange 2010 mailbox I also have two server that will host the CAS role in CAS ARRAy. I plan on using DAG on my mailbox server.

I have ready that you can setup a CAS Array using NLB which provide the distrubtuion of resource but not really  good for data redudance so I was thinking about using installing WIndows 2008 FailOVer clustiering to provide resouce and data redudancy?

Has anyone setup a CAS array using WIndows Failover clustering

ALos the cas array will also host the HUB transport role ..

0
compdigit44
Asked:
compdigit44
  • 6
  • 5
1 Solution
 
Praveen BalanSolution ArchitectCommented:
Hi Mate,

HUB role is by default fault tolerance(or would say load balanced). So you can install both Hub and Cas role in same server and configure NLB for CAS role.

Look at the links below
http://social.technet.microsoft.com/Forums/en-US/exchangesvravailabilityandisasterrecovery/thread/4fb99f9f-9d97-4432-9e14-195cd9783edf/

http://howtoexchange.wordpress.com/2009/12/16/configuring-client-access-array-for-exchange-2010-walkthrough/

-Praveen
0
 
NenadicCommented:
You will require more than one Mailbox server role to use DAG.  Since you already have three servers allocated - why not put all roles on all three?  That will provide you with a proper DAG and a very redundant transport layer.

Then, onto CAS.  Create a CAS array and then, depending on your userbase, pick up a hardware load balancer for $1500 from Kemp that will probably cover your requirements.

Back to NLB v. WFC.  CAS doesn't require data protection, so NLB is OK. However, it doesn't provide the best affinity options required for CAS.  Which is the reason a hardware load balancer is a better option.
0
 
compdigit44Author Commented:
More info.

I'm setting my my site and exchange 2010 enviroment for full folterance.

Site 1
2x physical servers in a ARRAY hoping for Failover cluster

1x mailbox server (primary)

Site 2

2x physical CAS servers in an array

1x mailbox server



All of the reasearch that I have does always talks about a CAS array on a NLB.

a hardware NLB is not an option for me.

CAn a CAS Array be setup on  a Failover cluster

BTW..

My current CAS nodes NIC are setup as follows

CAS1
NIC1, 2, 3 = teamed with static IP
Nic4 = statis IP connected via x-cable other cas1 for heartbeat

CAS2
Nic1,2,3 = teamed static ip
NIC4 = x-cable connected to cas1

I already created a static DNS record for my cluster NLB

It seems that a NLB in will used more static IP that an failover cluster

is this correct?
0
Making Bulk Changes to Active Directory

Watch this video to see how easy it is to make mass changes to Active Directory from an external text file without using complicated scripts.

 
NenadicCommented:
WFC cannot be used for CAS. Not properly, anyway. NLB will achieve what you need it to achieve, though. I'd suggest that you unteam your NICs, as there are issues with majority of software implementations of teaming.  Using a single NIC in each server will be more than enough. A single 1Gbps network connection is probably able to support a few thousand standard users anyway, before it becomes a bottleneck.
0
 
compdigit44Author Commented:
question

does my heat beat NIC connect need to be on the same LAN / network segement as the cluster virtual ip?

I cannot connect to my virtual cluster IP using the config listed above

I think it's a routing issue??

Suggestions
0
 
NenadicCommented:
Heartbeat is fine on its own segment, as long as the servers can communicate over it.

Did you try to unteam the NICs?
0
 
compdigit44Author Commented:
Yes didn't help

I need the teamed NIC to help handly all of my connections

I have read other users have had problems access the virtual cluster ip and had to added the default gateway to the heart beat network in or to get things to work

This doesn't make sence/. From my experience witth 2003 clustering the heart beat has nothing to do with the incoming LAN connections????
0
 
NenadicCommented:
How many connections do you have? As I mentioned, a single NIC on a CAS/HT server should comfortable handle a few thousand connections.
0
 
compdigit44Author Commented:
about 1500+ connections

I also want fault tolerance on the NIC incase it fails

all of my server use HP NIC teaming software
0
 
NenadicCommented:
Using teaming with clustered NICs can work, but it's not recommended. Similarly, based on those connections, you should reconsider hardware load balancing. Finally, based on what your connections are, you will have a single mailbox server in a site with 3000+ users. It's probably worth reconsidering your overall design and adding more fault tolerance on the mailbox server side.
0
 
compdigit44Author Commented:
Ok..

so let me make sure I'm understanding thing correcty..

1) You cannot use Teamed NIC's with NLB
2) Exchange 2010 does not support a CAS server installed on a FailOver Cluster

Correct?
0
 
NenadicCommented:
Hi compdigit44,

Close. The following are more correct statements:
1. NIC Teaming and NLB can have unpredictable results and is not recommended.
2. Exchange 2010 does support CAS server installed on a Failover Cluster, when failover cluster is used for DAG. However, it does not support WFC and NLB on the same computer.  Further, CAS has no use of Windows Failover Clustering. In other words, WFC cannot be used to provide load balancing for CAS, as it is a fault tolerant solution for different types of resources.

NIC Teaming and NLB:
http://support.microsoft.com/kb/278431

Regards,
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.

Join & Write a Comment

Featured Post

Free Tool: ZipGrep

ZipGrep is a utility that can list and search zip (.war, .ear, .jar, etc) archives for text patterns, without the need to extract the archive's contents.

One of a set of tools we're offering as a way to say thank you for being a part of the community.

  • 6
  • 5
Tackle projects and never again get stuck behind a technical roadblock.
Join Now