Multiple Mailbox & Hub Transport Servers Routing Issue (Full Version)

All Forums >> [Microsoft Exchange 2010] >> High Availability



Message


Exadmin143 -> Multiple Mailbox & Hub Transport Servers Routing Issue (19.Sep.2012 3:45:27 PM)

Dear All,

Before, I explain my issue, I must really appreciate all members on

this forum for their replies and help. Hats off to you all.

Well... I am a bit confused with my Exchange 2010 Organizational Level

Setup. We have 4 branches (more to come in future) namely Branch A,

Branch B, Branch C, Branch D respectively. Following is the

setup for my Exchange 2010.

Branch A,C & D in New York City
Branch B in Mexico


Domain Name: xyz.local
Branch A Primary DC - OXYGEN.xyz.local
Branch A Additional DC - HYDROGEN.xyz.local
Branch A Mailbox Server - XEON.xyz.local
Branch A HT+CAS Server - COBALT.xyz.local

Branch B Additional DC - BORON.xyz.local
Branch B HT+CAS+Mailbox Servers - BARIUM.xyz.local

Branch C Additional DC - NITROGEN.xyz.local
Branch C Mailbox Server - NEON.xyz.local
Branch C HT+CAS Server - CADMIUM.xyz.local

Branch D Additional DC - LITHIUM.xyz.local
Branch D Mailbox Server - RADON.xyz.local
Branch D HT+CAS Server - ZINC.xyz.local

Every respective user mailbox of each Branch, its mailbox resides on its respective Branch's Mailbox Server


All Branchs are interconnected with WAN links.

But, Specifically on Branch C which is recently configured with around 2000 users, I have observed for some reason the Microsoft Outlook Client while configuring the email account, despite providing the server as CADMIUM.xyz.local, automatically resolves to random Server i.e either BARIUM.xyz.local or ZINC.xyz.local.

Moreover, for some reason, when email sent from Branch C to Branch A, the messages pass through Branch B's Server i.e BARIUM.xyz.local, although Branch B is a far site and if VPN between Branch A & B is down, then messages sent to Branch A users from Branch C users bounce back with either Message Delayed notification. But sometimes, the message is either failed or Delayed after 24 hours.

I am really confused.

If Branch B VPN is down, then the users who are automatically got the Barium.xyz.local in their email client stop working. However, OWA works for them.

Issues need to be resolved :
1. Microsoft Outlook must be forced to accept the local HT+CAS server name and not random
2. If one branch VPN is down, its server must not be used to route the message even if the users 1 sending to user 2 both are in Branch C, but still message gets stuck in BARIUM.xyz.local queue. Strange !!!

3. Is it possible, I can force using Send Connector that messages intended to Branch A users must pass only through a specific connector.

Please note : That only Barium.xyz.local located in Branch B which has typical Exchange Setup with HT+CAS+Mailbox Roles installed. Rest are separate as mentioned above.
Branch D servers are ready but users not yet created, hence didn't test full mail flow from this site.
DAG Not yet configured.

Appreciate your suggestions !
Thanks




de.blackman -> RE: Multiple Mailbox & Hub Transport Servers Routing Issue (20.Sep.2012 8:40:11 AM)

Can you give us an idea of your Active Directory Site layout? Do you have a CAS Array configured for each of these branches, which I assume are in different AD sites? Have you configured the RPCClientAccess entry for the databases on each mailbox server to point to the CAS Array locally?

Look at your AD site configuration - remember that Exchange uses AD sites, site links and site link costs to determine mail routing so whatever path your AD takes to replicate will be the same for mail flow.




Exadmin143 -> RE: Multiple Mailbox & Hub Transport Servers Routing Issue (24.Sep.2012 11:48:37 AM)

Thanks for the reply de.blackman... sorry for the delay.. I wasn't feeling well. Anyways, I am ok now..

Its juss one AD with single forest and multiple Additional Domain Controllers (ADC) on each branch.

With your hint of RPCClientAccess pointing, I have found out that Mailbox servers of one particular branch is pointing to CAS of another branch which is as follows:

Get-MailboxDatabase | select name,rpcclientaccessserver | ft -auto

Mailbox DB Name                                RpcClientAccessServer
----                                                          ---------------------
System MBD01                                      COBALT.HMG.LOCAL
XE-MBD02                                            COBALT.HMG.LOCAL
XE-MBD03                                            COBALT.HMG.LOCAL
XE-MBD04                                            COBALT.HMG.LOCAL
BAR-MBD01                                         BARIUM.HMG.LOCAL
New-XE-MBD02                                   CADMIUM.HMG.LOCAL
New-XE-MBD03                                   CADMIUM.HMG.LOCAL
New-XE-MBD04                                   CADMIUM.HMG.LOCAL
NE-MBD01                                            CADMIUM.HMG.LOCAL
NE-MBD05                                            COBALT.HMG.LOCAL
NE-MBD03                                            BARIUM.HMG.LOCAL
NE-MBD04                                            ZINC.HMG.LOCAL
NE-MBD02                                            ZINC.HMG.LOCAL
-----------------------------------------------------------------------------
NE    - NEON (MBX Server) on Branch C
XE    - XEON (MBX Server ) on Branch A
BAR - BARIUM (HT+CAS+MB Server) on Branch B

BARIUM (MBX+HT+CAS) on Branch B
CADMIUM (CAS+HT) on Branch C
COBALT (CAS+HT) on Branch A
ZINC (CAS+HT) on Branch D

What I see from the RPCClientAccessServer Pointing to CAS Server on another site but instead it should be as follows:
XE-MBD must point to COBALT
NE-MBD must point to CADMIUM
BAR must point to BARIUM

could you plz confirm me the command to map them or point them to appropriate CAS servers.

Also, I believe that my AD is being replicated automatically. How do we replicate for proper mail flow.

Thanks again !!!




Exadmin143 -> RE: Multiple Mailbox & Hub Transport Servers Routing Issue (24.Sep.2012 11:55:36 AM)

I forgot to mention, there is no CAS Array configured... 




Page: [1]