CAS Array and DAG Site Resiliency (2 Remote Sites via Point to Point Protocol)

Hi, we are planning to deploy Exchange HA and Site Resiliency but i had some troubles understanding some concepts.
Out current configurations settings:
Datacenter A:
1 MB/CAS/HT (MB01)
1 CAS (CAS01)
1 Point to Point link between Datacenter A and Datacenter B (We create a VLAN in the remote datacenter pointing to the Active Directory site where our Exchange server is).
1 Active Directory Site with 2 Domain controllers (with VLAN 192.1.3.x in Datacenter A and VLAN 192.1.20.x in Datacenter B)
Datacenter B:
We want to install a new Exchange CAS Server (CAS02) in VLAN 192.1.20.x, is it posible to create a CAS Array with this configuration considering the replication network for the cluster in the MS Windows Server NLB, so that our Virtual IP of the cluster points
to 192.1.3.x VLAN? What are the network considerations to the replication and MAPI network between VLANs? Does we need some third party NLB or this could be achieved with MWS2008R2?
Same configuration for the DAG, with a new server (MB02) just for the MB role. The idea is to remove the CAS role from MB01 and have 2 Exchange Mailbox servers members of the DAG and 2 Exchange CAS Servers members of the CAS Array, so in case of a disaster
in Datacenter A our clients automatically failover to Datacenter B in this case all of our clients will be pointing to the CAS Array in their Outlook.
Hope you can help me with this, i don't know if this is possible with our actual infraestructure.
Best Regards,
Gerardo

Thanks Ed,
So in that case we will need at least 4 servers in Datacenter A (2 MB members of DAG and 2 CAS/HUB members of the CAS Array) to get a decent HA solution, Am I correct?
What about the other server in Datacenter B for DR, it will be a passive node or just another multirole server. So in case of DR (Let's say our Datacenter A is totally wrecked) I will need to switchover manually with my DB backups and switchover the user
mailbox and outlook clients to point to the DR server? I didn´t get the point of the DR server in Datacenter B, can you explain a little bit more?
I'm totally agree with you about the false positive failovers because our network doesn't meet that requeriments at all.
Hope you can help me to clear my mind.
Best Regards,
Gerardo

Similar Messages

Migration Exchange 2010 to Exchange 2013 with CAS Array and DAG

Dear All,
I am starting the migration of Exchange 2010 2 servers (CAS/Mailbox) with DAG no CAS Array to Exchange 2013 with 2 servers CAS array and 2 Mailbox servers with DAG. I read on some blogs that no requirement of CAS array on Exchange 2013. My concern how to
configure NLB on CAS servers for the client to connect.
Please guide and have any deployment guide for this, kindly share.
Thanks

Hi ,
As you said there is no use and meaning of having the cas array in exchange 2013 and also thanks a lot to Microsoft for introducing an single namespace facility in exchange 2013.
My suggestion and Microsoft recommendation should be to go with hardware load balancers for exchange 2013 rather than using the windows NLB and round robin method.
Why we need to go for HLB ?
Disadvantages
of some load balancing methods :
Windows
NLB :-
If you use Windows NLB then it can provide redundancy on server level failure and not on application level.
DNS
round robin :-
In case if we use the windows round robin method for load balancing then it wouldn't provide server level
and application level redundancy during the failures.At the Same time we need to manually adjust the DNS records during the server failure but on the client end dns caches will create the issues.
Configuring NLB for exchange 2013 :
http://msexchangeguru.com/2013/08/14/windowsnlb/
NLB configurations for exchange 2010 and 2013 will be same.
Configuring round robin for exchange 2013 :
http://exchangeserverpro.com/exchange-2013-client-access-server-high-availability/
Advisable method is to have the CAS and MBX roles on the same box if NLB not comes it to play.Because windows failover clustering and NLB cannot be configured on the same box.In exchange 2013 cas role is a stateless server role so there is no need to have
that role on a separate box.
Thanks & Regards S.Nithyanandham

Planning for CAS Array with DAG intergation on Exchange 2010

We currently have 2 Exchange servers and have plans to add 2 more servers so it will be 2 servers with DAG and 2 servers with CAS array
Here is our current setup: EX01 has HUB, CAS and MAILBOX roles EX02 has MAILBOX only with DAG configured between 2 servers.
Should I install another Exchange (EX03) with Mailbox role and remove it from EX01 and then have another EX04 to setup CAS array? What is the best practice ?
alex serdyukov

Hi,
In your case, you can install another Exchange (EX03) with Mailbox role and remove Mailbox role from EX01 and then have another server EX04 to setup CAS array.
Then you can deploy DAG using server EX02 and EX03, and add server EX01 and EX04 to CAS array.
Best regards,
Belinda Ma
TechNet Community Support

CAS Array and Outlook anywhere profiles

Hello !
I'm planning to set up a CAS array from an existing CAS server. I red that after changing RpcClientAccessServer parameter, users Outlook profiles would not be updated automatically for internal users, but I was wondering if it could work for external users
using Outlook Anywhere ? We have no internal user (since our Exchange environment is in a datacenter).
Thanks

Hi,
After changing the RpcClientAccessServer parameter, all clients including Outlook Anywhere users will lose connection without any configuration update.
For more information, you can refer to the following article:
http://blogs.technet.com/b/aljackie/archive/2013/11/14/outlook-rpc-end-point-and-pf-the-microsoft-exchange-administrator-has-made-a-change-that-requires-you-quit-and-restart-outlook.aspx
“To add further on the CAS array, the Outlook clients with an existing Outlook profile would continue to use the old RPC endpoint rather than the new RPC endpoint (even though Autodiscover detected the change). This is because the old RPC endpoint does not
return an ecWrongServer response to the client. The RPC endpoint accepts the connection; therefore, Outlook ignores the Autodiscover response because it has a working connection.”
We can recreate profile or move database to resolve the issue.
Best regards,
Angela Shi
TechNet Community Support

Two dag member no cas array

I have two exchange servers (exchange 2010) in a dag with two databases. The production exchange server and a dr exchange server. Both of the exchange servers have a cas and I have not setup a cas array. I have set the -RpcClientAccessServer to the
production exchange server.
Is this a valid configuration?
Is it possible to prevent connections to the second exchange server? I find on email headers and on outlook 2010 connection status that there is connections to the second exchange server constantly, even though all mailboxes are on the production server
and -RpcClientAccessServer is production server.

I will look at creating a CAS array then. I thought i would be able to get away with out it. I think what happened was that after adding the second exchange server i did not update the rpclientaccesserver property to point back to the production server,
for some reason this automatically changed itself to the new exchange server. This means for a few weeks the users were connecting to the DR exchange server over a slow 100mbit link. As soon as i realized that i changed the rpcclientaccesserver property back
to the production CAS on both databases.
I found this http://www.mikepfeiffer.net/2011/04/determine-the-number-of-active-users-on-exchange-2010-client-access-servers-with-powershell/
Very handy powershell script that tells you client connection status to the CAS. currently we have 85 users on production CAS and 20 still on the DR cas even though i set it back to correct CAS over a week ago. Now i need to look at moving these users over
to the correct cas. But thinking might as well wait till we set up CAS array and then change all users that way.
the reason i dont have active active is because its over a 100mbit link and is not as fast as the gigabit we have to the production exchange and is not on as fast storage or hardware. The DR exchange just sits on local storage on one esxi box with one other
server. The rest of our vms are snap mirror across with netapp snap mirror to third party datacenter. In the event of a dr they will supply some new servers for use with the snap mirrored vms.
In the event of a DR situation there is the understanding that it may take a day before we have everything up and running and this is accepted by the client. We are looking at improving this though and would be interested to hear any suggestions regarding
improving exchange dr.
I had some problems with the DAG last month. I had problem seeding one database and finally managed to get that workign and recently had to move the witness to another server as that stopped working with strange error. But now the DAG is 100% healthy with
no issues, i think the 20 random email problem was caused by stuck emails in the users outlook 2010 on xenapp that he never reported, which is another known issue when migrating from outlook 2003-10. Still very concerning however.
It would be very handy if there was some automated way to force outlook to the correct CAS only, say by changing a registry key with group policy. If anyone knows of any way to do that, would be helpful thanks.

Creating a cas array for exising prd mailbox servers

Hi
one of the production site in current environment , mbx databases rpcclientaccess server set as individual cas servers .
we want to point these databases to a cas array ,NLB is already created now remaining is cas array and point the database to cassarry fqdn.
I just want to know when we do this change , any client re-configuration is required or automatically redirection will happen to cas arary from outlook client.
Regards

I just want to know when we do this change , any client re-configuration is required or automatically redirection will happen to cas arary from outlook client.
Hi,
I'm afraid that you need to manual re-configuration from outlook client.
I recommend you refer to the following article:
Demystifying the CAS Array Object - Part 2
5.A CAS array object should not be configured after creating Exchange Server 2010 databases
The profile will not update itself because the client will not receive an
ecWrongServer response from CAS. It will not receive this response because any CAS is a valid connection point for any mailbox database via RPC (over TCP) so clients can survive datacenter switchover/failover events without being reconfigured and all
an admin has to do is flip the CAS array object DNS record to point to a surviving pool of CAS. Currently the only way to fix mailbox profiles would be a manual profile repair within Outlook, by publishing an Office PRF file via GPO (not going to work for
non-domain joined machines), or by decommissioning the CAS server named in the users’ profiles so the endpoint is no longer available. This last option should (test test test!!) trigger a full profile repair by Autodiscover in Outlook 2007 or Outlook 2010.
Outlook 2003 is only repairable with a profile repair or a PRF file. Autodiscover will not as of this article’s writing update a profile to a new server name as part of the normal Autodiscover process which updates the Outlook Anywhere configuration and discovers
EWS URLs for other features such as OOF Management, Free/Busy, and Inbox Rules management.
Hope this helps!
Thanks.
Niko Cheng
TechNet Community Support

CAS Array created but Outlook 2007 still points to first CAS Server

Hi All,
I just created a DAG between two servers and CAS Array between two other servers.
DAG:- EX01 -Mailbox, CAS,HT (First Exchange Server)
EX02-Mailbox
CAS Array: CA-HT01 (CAS & HT Role)
CA-HT02 (CAS & HT Role)
I created CAS Array and assigned an FQDN for that array but It does not get changed in outlook 2007 it still uses old server EX01.So when I am shutting down server ex01 for testing all outlook 2007 client
get disconnected.
Does this required to change profile for all outlook 2007 users? I am more than 1600 users I am not sure how many users using outlook 2007.
Please guide.I already configure Autodiscover service as it is suggested somewhere online but no luck.
Thanks

Mike,
I tried with shutdown original RPCCLientAccessServer for two days but still that old value did not
chnaged to new one.Can you guide me what is the exact method of creating Autodiscover so that outlook 2007 will point to new RPCCLientAccessServe value when
existing is down/not accessible.
Thanks
If the RPCCLientClientAccess Server referenced in the Outlook profile is not accessible, then Outlook should have found the new one at some point or failed to connect to Exchange entirely. Are you saying the Outlook clients continue to connect to Exchange
even though the server defined in Outlook was not accessible?
If you close and re-open Outlook does it show the new RPCCLientAccessServer value in the profile?
Twitter!: Please Note: My Posts are provided “AS IS” without warranty of any kind, either expressed or implied.

Exchange 2013 mailbox added to the CAS array

We are upgrading to Exchange 2013 from Exchange 2010. Following the development guide, we have
installed the first mailbox server in the Exchange 2010 environment which has 3 Exchange 2010 CAS server
in the array. When installing the mailbox role, we did not choose the client access role but after
the installation we can see that the Exchange 2013 Mailbox server is added to the CAS array and yet
we did not choose the client access role. How does this happen, and to proceed ?

I see the same thing in my lab:
Get-ClientAccessArray | FL
RunspaceId        : 16b992b3-270f-4ae1-a3c3-fa9e2ea73d69
ExchangeLegacyDN : /o=Wingtiptoys/ou=Exchange Administrative Group (FYDIBOHF23SPDLT)/cn=Configuration/cn=Servers/cn=Ou
                    tlook.wingtiptoys.ca
Fqdn              : Outlook.wingtiptoys.ca
Site              : wingtiptoys.ca/Configuration/Sites/Default-First-Site-Name
SiteName          : Default-First-Site-Name
Members           : {EXCH-2010, EXCH-2013}
AdminDisplayName :
ExchangeVersion   : 0.1 (8.0.535.0)
Name              : Outlook.wingtiptoys.ca
DistinguishedName : CN=Outlook.wingtiptoys.ca,CN=Arrays,CN=Exchange Administrative Group (FYDIBOHF23SPDLT),CN=Administr
                    ative Groups,CN=Wingtiptoys,CN=Microsoft Exchange,CN=Services,CN=Configuration,DC=wingtiptoys,DC=ca
Identity          : Outlook.wingtiptoys.ca
Guid              : 27968af1-1624-4ff3-85c8-e38e68183afe
ObjectCategory    : wingtiptoys.ca/Configuration/Schema/ms-Exch-Client-Access-Array-2
ObjectClass       : {top, server, msExchExchangeServer, msExchClientAccessArray}
WhenChanged       : 4/12/2014 12:51:18 PM
WhenCreated       : 4/12/2014 12:51:18 PM
WhenChangedUTC    : 4/12/2014 7:51:18 PM
WhenCreatedUTC    : 4/12/2014 7:51:18 PM
OrganizationId    :
OriginatingServer : DC-1.wingtiptoys.ca
IsValid           : True
Cheers,
Rhoderick
Microsoft Senior Exchange PFE
Blog:
http://blogs.technet.com/rmilne
Twitter:   LinkedIn:
  Facebook:
  XING:
Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

CAS array internal DNS IP address best practice

Hi, Just a question about a best practice approach for DNS and CAS arrays.
I have an Exchange 2010 Org. I have two CAS/HUB servers and two MBX servers. My external DNS (mail.mycompany.biz) host record points to a public IP address which is NAT'd to the internal IP address of my NLB CAS cluster. I maintain a split brain
DNS. Should the internal DNS entry for mail.mycompany.biz also point to the public IP address or should it point to the internal IP address of the NLB cluster?

A few comments:
The reason you have split DNS is to do exactly these sort of things: inside users hit the inside IP and outside users hit the outside IP. You'll have to look at your overall network design to see if it makes sense for users to take this shortest route
to the services, or if there is value in knowing all users simply take the same path.
You should not be using the same DNS name for your web services (e.g. OWA) as you are for your CAS array. This can cause very long connection delays on Outlook clients, not to mention overall confusion in your design. Many orgs will use something
like "outlook.domain.com" for the Client Access Array and "mail.domain.com" for the web services. Only the later of these two need to be exposed to the internet.
Keep in mind, Exchange 2013 dramatically changes this guidance. There is no more CAS array, and the
recommended design is to use dedicated namespaces for each web service.
Mike Crowley | MVP
My Blog --
Planet Technologies

Outlook lost connection when load balanced CAS array name was moved to different AD site and data centre

At the weekend our F5 load balanced CAS array name was moved to our 3rd datacentre and all users lost access to Outlook. The databases were running on DAG2 listed below at the time.
We have 3 DAG servers configured as the following:
DAG SERVER ONE (DAG1) | Datacentre 1 | Primary AD Site | Multi role server MBX, CAS, HT
DAG SERVER TWO (DAG2) | Datacentre 2 | Primary AD Site | Multi role server MBX, CAS, HT
DAG SERVER THREE (DAG3) | Datacentre 3 | Secondary AD Site | Multi role server MBX, CAS, HT
We primarlily run from DAG1 or DAG2 - DAG3 is for real DR and loss of Datacentre 1 and 2. The CAS array "Outlook" is configured for the primary AD site load balanced across DAG1 and DAG2.
We use an F5 load balancer that balances the connections across DAG1 and DAG2, if both fail then DAG3 is used automatically. We have tested and it works fine simulating loosing DAG1 and DAG2 and bringing up as a 1 node cluster on DAG3 with a
FSW and the load balancer pointing at DAG3 for Outlook connections.
For some reason this stopped Outlook users connecting - can anyone advise me why as I have been asked to test this again and I was expecting it to work?

Hi,
As far as I know, For Exchange 2010 before SP2 RU3,the Outlook clients with an existing Outlook profile would continue to use the old RPC endpoint rather than the new RPC endpoint (even though Autodiscover detected the change). Thus, except with Andy's suggestion,
I recommend you can also repair or recreate the Outlook profile to have a test.
For more information, you can refer to the following article:
http://blogs.technet.com/b/aljackie/archive/2013/11/14/outlook-rpc-end-point-and-pf-the-microsoft-exchange-administrator-has-made-a-change-that-requires-you-quit-and-restart-outlook.aspx
Thanks,
Angela Shi
TechNet Community Support

Can CAS array include server in DR site

Hi, there:
Our current Exchange 2010 environment:
1): Two HUB/CAS Exchange 2010 server running on Windows 2008 R2 use NLB to form a CAS array.
2): Two Exchange 2010 Mailbox server with one DAG.
Please note all of these Exchange 2010 servers are on our datacenter, now we would like to add the DR capability to our Exchange environment.
In regards to the CAS array design for the DR site, is it possible to add one HUB/CAS Exchange 2010 to current CAS/HUB array (this server need to be on SR site), this way when the two HUB/CAS server on the datacentre is not available user can still connect
the HUB/CAS server on the DR site?
If the subnet for the Datacentre and DR site is different is this possible?
If the subnet for the datacenter and DR site is the same is this possible?

Hi,
Firstly, I’d like to explain, CAS array’s members cannot be located in different sites. If all CAS servers are in the same site, CAS array can load balance MAPI request.
And if all mailbox severs are in the same site, DAG can apply failover while it may apply switchover if they are in different sites:
http://technet.microsoft.com/en-us/library/dd298067(v=exchg.141).aspx
A failover is an automatic activation process that can occur at either the database or server level. Switchover is manual activation process.
If you have any question, please feel free to let me know.
Thanks,
Angela Shi
TechNet Community Support

Same CAS Array Exchange 2010 (HLB), with OS Windows 2008R2 and Windows 2012.

Hello,
We have a 10 node DAG (Exchange 2010 SP3, Windows 2008 R2), with 2 CasArray.
We are planning to add news (multirole) servers and create a new DAG (Exchange 2010 SP3, Windows 2012) in this infra, in the same AD site, to migrate all mailbox from the other DAG (Migration from virtual servers to physical servers).
So we use the same CasArray (HLB, with F5) with différent OS version, during the migration time (1 month or more). I haven't found anything that say it's not supported or can be problematic.
Have you feedback or advice?
Thanks,
Sébastien

Hi,
Based on my knowledge, there is no need to deploy a CAS array with CAS servers running on the same Window version. The version can be different.
After a Client Access server array is defined within an Active Directory site, all Client Access servers within that Active Directory site are automatically part of the Client Access server array.
Best regards,
Belinda
Belinda Ma
TechNet Community Support

Outlook Profile Not Updated to New CAS Array After Mailbox Move Between Sites

I am working on a large mailbox migration from one AD site to another.
Exchange version at both sites:Exchange 2010 SP2
Two Mailbox Servers: MBX01 in site A / MBX02 in site B
(There is a DAG but keeping this simple)
Two CAS Arrays: CASA.domain.com / CASB.domain.com
The RPCClientAccessServer property is set accordingly. Where MBX01 is set to CASA.domain.com and MBX02 is set to CASB.domain.com
Clients:Outlook 2010 SP1
Scenario: I successfully move a mailbox from Site A to Site B. The Outlook client does not automatically update its server settings to point to the new cas array (CASB.domain.com) and continues to connect to the old CAS server (CASA.domain.com).
I have seen a few articles regarding this pointing to workarounds such as:
- repairing the outlook profile on the client
- delete the existing outlook profile
- run a script to update the prf file (re-cache's email at the client)
I am interested in knowing if anyone else has experienced this, what their solution was for a large migration and if any roll-ups or service packs have possibly fixed this problem.
I would also like to know if you have seen this affect ActiveSync devices. I am getting mixed results in my tests and thought it may be contributed toward different device types, mobile os versions, etc...
I have posted this to the Outlook forum as well, just wasn't sure where to start.
Thanks.
Mike

Hi,
Firstly, I’d like to say, in Exchange 2010 and prior to SP2 CU3 version,
when we move mailboxes between AD sites, users will not receive any notice about restarting Outlook and their RPC endpoint won’t update to reflect the RPC Client Access Server array associated with the mailbox database in the AD site where the mailbox now resides.
Depending on repairing or recreating profile, we can force Outlook to use the new RPC endpoint.
If you have any question, please feel free to let me know.
Thanks,
Angela Shi
TechNet Community Support

CAS ARRAY fail-over and emails stuck

Dear all,
for some reasons we are in Exchange Server Coexistance mode that is 1 Exchange 2003 Server and 2. Exchange 2010 servers.
We have CAS array(node-1 And Node-2) and DAG in place , but the problem is whenever my Node-1 is down emails are getting stuck routing group connector on legacy server and Exchange 2010 to Exchange 2010 emails are working.
oppositly
when my Node-2 is down everything works.
how do I FIX this ?
TheAtulA

Dear all,
for some reasons we are in Exchange Server Coexistance mode that is 1 Exchange 2003 Server and 2. Exchange 2010 servers.
We have CAS array(node-1 And Node-2) and DAG in place , but the problem is whenever my Node-1 is down emails are getting stuck routing group connector on legacy server and Exchange 2010 to Exchange 2010 emails are working.
oppositly
when my Node-2 is down everything works.
how do I FIX this ?
TheAtulA
I assume the CAS also have the hub trasnport installed?
Check the Routing Group Connector(s) (Get-RoutingGroupConnector) and ensure the source and destination transports include both CAS nodes, not just Node-1.
If not, then use set-routinggroupconnector to set the correct source and target servers
https://technet.microsoft.com/en-us/library/aa998581(v=exchg.141).aspx
Twitter!: Please Note: My Posts are provided “AS IS” without warranty of any kind, either expressed or implied.

Exchange 2010 server (Mail,HUB and CAS) installation in new site.

We are planned to install Exchange 2010 servers(MAIL,HUB and CAS) server in new Active director site in existing exchange organization.
Kindly share best practice link for exchange server installation in new site.

The best practice for installing into a new Active Directory site is the same as installing into a new organization, except you don't need to prepare the organization or domains (unless you are also installing into a new domain that hasn't been prepared).
Install the CAS then hub, then mailbox roles, or install a multi-role server, then define your CAS Array name and IP address, and set your CAS behind a load balancer (if needed). Add your external certificate to both the CAS and the load balancer (and the
hub if you use TLS for mail transfer), and you should be golden. Also, if you intend to proxy your CAS connections, here's a link that may help:
https://technet.microsoft.com/en-us/library/bb310763.aspx?f=255&MSPPError=-2147217396
For completeness, here's a thread on the msexchange.org forums that specifically mentions adding Exchange into a second site:
http://forums.msexchange.org/New_Exchange_2010_at_another_Site/m_1800557445/tm.htm

CAS Array and DAG Site Resiliency (2 Remote Sites via Point to Point Protocol)

Similar Messages

Maybe you are looking for