Exchange 2010 WAN redundancy

Hi all, I am looking to add some redundancy for inbound/outbound SMTP. I have Exchange 2010 running all roles in one VMware VM. I was having trouble getting this to work with our firewall, so I spoke to tech support at sonicwall and their recommendation
was this.
Add a NIC in the Exchange VM for each WAN link that we have (Currently 2). Then use the sonicwall's Route Policies to connect each NIC to each WAN link (one NIC for one WAN) to use for outbound SMTP.
Does this sound like a good plan? Any pitfalls/issues I should be aware of?

OK great thanks. I am posting in both forums. In this context I was mainly curious if Exchange would have any problems with multiple NICs. In case of failure, should I expect it to try every available NIC when routing outbound
SMTP? Do I need to configure anything for it to be aware of each connection?
That's really a Windows question :) You can associate receive connectors with IP addresses, but not a send connector. Also, with multiple NICs, only one of them is going to have a defined gateway. Seems to me that the redundancy would be built in the
VMHost and/or switch/router etc... and not into the NICs on the Exchange guest.
Twitter!: Please Note: My Posts are provided “AS IS” without warranty of any kind, either expressed or implied.

Similar Messages

Exchange 2010 DAG's - 4 Mailbox Servers in 2 Datacenters with per site redundancy

I recently came into a company with an Exchange 2010 sp1 environment where there are two data centers in two different subnets (10.70.62.0 and 10.80.56.0)
Each datacenter has 2 mailbox servers (4 servers total) but their plan was to have only enough storage space to host 1/2 of the databases on each server. So instead of server 1, 2, and 3 being copies of each other, ServerA is a copy of ServerA and ServerB
is a copy of ServerB
Server     -    Databases
Site1mbA - Primary Database copy 1-10
Site1mbB - Primary Database copy 11-20
Site2mbA - Secondary Database copy 1-10
Site2mbB - Secondary Database copy 11-20
Normally if there were enough storage space I would keep a copy on each server and replicate within 1 DAG.
However since not all the databases are on each member server do I need to use 2 DAGS? 1st for Site1mbA & Site2mbA and 2nd for Site1mbB & Site2mbB
I feel I would need the 2 DAG's because with 1 DAG if Site1mbA server went offline the DAG couldn't switch to the Failover Cluster IP for Site2 because there are still active mailboxes in Site1 on Site1mbB.
Does that make any sense or am I over complicating it?
Thanks!

There is no rule that every server in a DAG must have a copy of every Database. You could do this with one DAG, but if the WAN goes down then the DAG will be active in only a single site not both.
http://technet.microsoft.com/en-us/library/dd979781(v=exchg.141).aspx#FourTwo
As illustrated in the previous example, using a single four-member DAG extended across two datacenters can provide high availability and site resilience for the mailbox services and data. However, if a WAN outage occurs, only the primary datacenter retains
service because it contains the majority of the voters. The datacenter with the minority of voters loses majority, and the DAG members in that datacenter lose quorum and go offline.
To deploy highly available Mailbox servers in a multiple datacenter environment, where each datacenter is actively serving a local user population, we recommend that you deploy multiple DAGs, where each DAG has a majority of voters in a different datacenter,
as illustrated in the following figure.
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.

Exchange 2010 - Method to send e-mail when the WAN IP is dynamic

Hello,
I am writing due to the fact that I cannot send e-mail from my Exchange 2010 Server to my Yahoo! address for testing. When trying to do so, I get the following error message:
mta1106.mail.gq1.yahoo.com gave this error:
Mail from xxx.xxx.xxx.xxx not allowed - 5.7.1 [BL21] Connections not accepted from IP addresses on Spamhaus PBL; see
http://postmaster.yahoo.com/errors/550-bl21.html [550]
As one can see, because my IP address is residential, I cannot send e-mail to all addresses on the web. I really would like to use this as my primary contact, and thought of using a proxy. Can someone recommend a good proxy, or workaround for this? I am
sorry for posting in Exchange 2013, but I couldn't find the 2010 forum. Thank you to all who reply.
Mike

Thank you for your reply,
I have set up DynDNS, and I am currently using my own domain name. The problem was I needed to configure a smart SMTP host, or relay server on my send connector. I decided to go with SMTP2GO, as they are cheap, and allow you to send 20 e-mails a day for
free. When the money is right, I plan to upgrade.
Thank you for your help.
Mike

Upgrading from Exchange 2003 Bridgehead servers to Exchange 2010 Transport Hub servers for routing SMTP only

Our company moved from on-premises Exchange 2003 to Office365 and only have 4 Exchange 2003 servers on-prem that we use for Routing email from application servers to Office365. We need to migrate these servers to Exchange 2010 then to Exchange 2013 and
only route email only. Is it possible to upgrade to Exchange 2010 by installing the transport Hub & Mailbox server options only? Our OAB and EWS services come from the CAS servers located on Office365 so we should not need a CAS server to set up Transport
rules or route mail would we? Any assistance with this would be greatly appreciated! - Thanks, DWB
Dave

Since it is not internet facing we will not have to worry about configuring an Internet Domain Name when installing the first server, correct? in my planning I was going to
install the Exchange 2010 CAS/Mailbox/Transport Hub roles on one server, then upgrade it to Exchange 2013. Once this is completed I'll then install the 2 mailbox role servers in one Datacenter, and 3 more in our DRP DC. If something happens to the
primary Datacenter I would want it to fail over to the other site. For this I would have to install another CAS/mailbox server in the secondary DC. Would this plan sound about right? Since we moved to Office365 in 2010 I have not had a chance to deal with
actual servers except for the Exchange 2003 servers we still have on-premises. Each of these are located in 4 Regional offices along with single Windows 2008 R2 servers using only the IIS SMTP service for routing mail from on-prem application servers
and print/scanners which email back to the users. The plan is to move to a routing system which will provide both MTA and redundancy if one has an issue.
Dave

Exchange 2010, Outlook 2010, some clients receive "server unavailable" when logging into outlook

We have a single exchange 2010 server running hub, cas, and mailbox as a virtual machine in VM 5. Currently on version 14.3 build 123.4 running on Server 2008 R2. Clients connecting to this server are all running Outlook 2010 with a mix of OS between
xp and win 7. Domain controllers are all 2003 sp2. We have a multi-campus environment with most users accessing across a WAN connection (mpls and metro-e). Quota Limits of 1gig imposed on all mailboxes. We do still
have a Exchange 2003 server that we are completing the migration of users off of in our environment. But all mailboxes on exchange 2010 are accessed thru Outlook 2010. Migration from 2003 to 2010 happens only after Office 2010 is installed.
We are a little over 50% migrated - (~750 users, 1100 entities)
I have some 2010 clients complaining that they receive "the server is unavailable" when launching outlook for the first time in the morning. A second time it usually opens. This happens for remote WAN and local LAN users. Also,
during the day as outlook is open, when trying to open an email or navigate between folders in the inbox they will also have outlook stop responding as it tries to contact the server.
Monitoring the server - CPU averages 40-50% thru the day with spikes in the 75-100%. Usually the processes that spike are Store.exe or msftefd.exe.
I have verified that IPV6 is disabled on the exchange server - and it doesn't exist on the 2003 DC's. Searching for answers keeps turning up results for OOF or Outlook 2003 RPC. Neither are the case here.
Would we benefit from adding a 2nd server into our environment? If so - hub? Cas? which would be better to offload off the primary?
or is there a better way to handle this issue?

Hi,
According to your description, the issue happens on every morning. Thus, it’s mostly likely caused by that there are too many sessions bringing the performance issue.
To confirm it, we can use performance monitor to check the following counters when Outlook works slowly:
RPC request, RPC average latency and RPC operation/sec
http://technet.microsoft.com/en-us/library/dd335215(v=exchg.141).aspx
If you have any question, please feel free to let me know.
thanks,
Angela Shi
TechNet Community Support

Outlook / Exchange 2010 shared calendars performance

Hi, we are using Outlook 2010 connecting to Exchange 2010 over a WAN (120ms average RTT). Cached mode is forced on all clients. Our users are finding that performance of shared calendars is very poor in "online" mode. However as they are all using Outlook
2010, checking "download shared folders" means that ALL secondary mailbox folders are cached and we don't want this to happen. Users are typically PAs who have been granted full mailbox access to their exec's mailbox. These users may have several shared calendars
open at once which makes the situation even worse.
Is there a way that we can revert Outlook 2010 to the previous 2007 behaviour whereby with secondary mailboxes only the calendar / contacts etc are cached? I have seen articles referring to a reg value "CacheOthersMail" that can be added but this seems to
have no effect on a couple of clients we've tried - if the "download shared folders" option is checked, everything is still cached, and if it's unchecked, everything switches to "online".
Any help appreciated,
Nick

Yes, setting that registry key to 0 should do the trick. You might need to recreate the mail profile though or recreate the ost-file to directly get rid of what is currently being cached.
As you've currently forced the caching by a GPO, check your policy keys as well.
See; http://www.msoutlook.info/question/561
Robert Sparnaaij
[MVP-Outlook]
Outlook guides and more: HowTo-Outlook.com
Outlook Quick Tips: MSOutlook.info

Introduce an "passive" Exchange 2013 in Exchange 2010 environment without user or administration impact.

Hi
We wish to introduce an Exchange 2013 server in our existing Exchange 2010 environment.
The Exchange 2013 server is only needed for migration purposes where we wish to leverage the new features of the migration-batch functionality.
It’s important that the Exchange 2013 server will not have any “visible” impact on the Exchange 2010 environment, or take over any functionality, right now, since the customer wish to stay on 2010.
Can you guys come up with anything to be aware of?
Oblivious we disable the Autodiscover SCP for the Exchange 2013 server, so that the clients will not hit this server. Nor will we point any URLs to the Exchange 2013
I can see that a new “Default Offline Address Book (Exch2013)” is created and set as default when EX2013 is installed. We will change this back to the default EX2010.
The server will properly take part in the Shadow Redundancy feature, and is doesn’t seem to be possible to exclude the server or avoid, unless disabling Shadow Reduncancy.
I can only come up with these 3 things that will have a "direct" impact on the environment, but any input will be highly appreciated!
BR,
Martin

Hi,
I agree with Li Zhen’s suggestion. If you don’t migrate from Exchange 2010 to Exchange 2013, we can disable add services in Exchange 2013 and don’t use it any more.
If you want that the Exchange 2013 server will not have any “visible” impact on the Exchange 2010 environment, or take over any functionality, please keep the published server to pointed to Exchange 2010. Then the original configuration in Exchange 2010
would not be changed to the new configuration in Exchange 2013.
If you want to use Exchange 2013, then we can point the published server to Exchange 2013 and configure the virtual directories in Exchange 2013. Though all mailboxes are still located in Exchange 2010, all external requests would be proxy or redirected
from CAS 2013 TO CAS 2010 automatically.
For more information about Client Connectivity in an Exchange 2013 Coexistence Environment, please refer to:
http://blogs.technet.com/b/exchange/archive/2014/03/12/client-connectivity-in-an-exchange-2013-coexistence-environment.aspx
Regards,
Winnie Liang
TechNet Community Support

Hardware requirement for running exchange 2010 on vmware workstation

Dear Team,I need to create a exchange 2010 sp2 on a guest os on vmware workstation.kindly help me out with the physical hardware requirements for a home user

Hey Mukund26,
Before I answer your questions, my suggestion would be for you to figure out exactly what you want to get out of running that Exchange lab. Are you trying to learn Exchange features like DAGs? Are you trying to learn about virtualizing Exchange? When you figure that out you'll be better able to answer question 1 since you may not need all of those virtual machines.
Here are my thoughts on your questions.
1) It is probably technically possible to run 10 virtual machines on the configuration you have but you're going to have pretty bad performance. I would try to trim that down to only the virutal machines that you really need in order to use the lab how you want to use it. Here are some questions that may help you figure out what you need:
Why do you need two domain controllers? I don't think you need to worry as much about redundancy in your lab.
Why do you need 2 ESXi hosts and a vCenter? What purpose do they serve in this configuration? Are you trying to run the Exchange VMs nested under the ESXi VMs? If so, why?
Is there a reason why you want to separate out HTS/CAS from MBX in your lab? Combining all roles together can reduce the number of VMs you need. If you separate out the roles you'll need to allocate a lot of RAM to support all of those Exchange VMs.
2) Running a lot of virtual machines on a single 1TB SATA disk, especially a 5400 RPM disk, is going to be slow. It'll be especially painful when the VMs are booting up, rebooting, shutting down, etc.. Try to stagger the power on operations so they don't all boot at once. For example, first boot your DCs, then boot your HTS/CAS, then your MBX, etc. There is no doubt that configuration will be slow, so if you can get your hands on an solid state drive with enough space you'll have significantly better performance. For the purposes of a lab, though, it is probably ok if you can be patient.
3) I think the i5 processor will be just fine for this lab. Usually in a lab setting CPU performance isn't the most important thing so similar to above as long as you can be a little patient I think that CPU will suit you just fine.
As I said figure out what you want to get out of the lab and then you can determine how many VMs you need and whether or not you want to upgrade hardware. If you're looking to learn about Exchange 2010 in general I would suggest using multirole servers, getting rid of one of the DCs, and not worrying about virtualizing ESX/vCenter so that you can learn. If you're looking to learn about virtualizing Exchange 2010 on vSphere then let's figure out exactly what you want to get out of it and figure out what you can learn using VMware Workstation and what you really need vSphere for.
Best of luck!
Matt
http://www.thelowercasew.com

Database redundansy health check failed ( Exchange 2010 sp3 )

Hello all ,
One of the mailbox server have been getting the following error in in application , sometimes
RealCopyQueu is 256 and sometimes 0 and healthy .
Kindly share how can resolve it please

Hi,
I suggest we refer to the following article to troubleshoot this issue. Run
Database Redundancy Check Script.
http://social.technet.microsoft.com/Forums/exchange/en-US/b1cfbc4b-b17d-4148-adce-1a565d0f5746/exchange-2010-event-id-4113-msexchangerepl-health-check-failed
Regards,
Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact
[email protected]
Simon Wu
TechNet Community Support

Multiple roles load balancing on Ms Exchange 2010

Dear list member !
Currently. I installed Ms Exchange SP3 Multiple roles on a single server. I have been planning deploy add a member exchange 2010 SP3 for redundancy DAG but these are
people also IT operation told to me so install CAS, HUB, MB Roles will Load balancing CAS, DAG based on TWO Server "Ex 2010 SP3:". Following Microsoft document, then almost do not that.<o:p></o:p>
Please feedback to able or unable
very appreciate

Hi,
To load balance CAS servers, you need to use the Windows Network Load Balancing or the Hardware Load Balancing. For more details about this, you can refer to the following article.
Understanding Load Balancing in Exchange 2010
http://technet.microsoft.com/en-gb/library/ff625247(v=exchg.141).aspx#options
If you want to deploy DAG in your environment and you also want to load balance your CAS servers, it is recommended to install Mailbox server role and CAS role on different servers. Because DAG members utilize Windows Failover Clustering, which can’t co-exist
with WNLB. Of course, you can choose to use HLB to load balance CAS servers.
Best regards,
Belinda
Belinda Ma
TechNet Community Support

Exchange 2010 to 2013 Migration- orphaned 2010 servers in Edge subscription

Migration of Exchange 2010 to 2013 was carried out. Steps to decommission the 2010 servers was done, including removing the Exchange software fro the servers. Looks like some pieces were missed though as the 2010 servers still appear in the smarthostdelivery
and shadow redundancy on the 2010 edge transport server.
How can these be removed?

Hi,
I recommend you can use the following script to cleanup the queue:
Cleanup Shadow Redundancy Queues
Performs a search per server per shadow redundancy queue. If shadow redundancy queues contain
at least 1 message the receive date is inspected. If it is more then 1 day ago it will cleanup
the message.
Normally you do not have to cleanup the shadow redundancy queues because messages will automatically
expire, how long this takes depends on your config but normally 2 days.
However in some scenarios it might be necessary to clean the queues.
Hope this helps!
Thanks.
Niko Cheng
TechNet Community Support

Moving Exchange 2010 Production DAG Setup and Servers to difrent Date Center with new IP Addresses

My Setup
Exchange 2010 SP3 2 Mailbox, 2 Hub/CAS, 2 TMG Servers in Production
Exchange 2010 SP3 2 Mailbox, 2 Hub/CAS, 2 TMG Servers in DR Site
Exchange 2010 DAG implemented with site level redundancy
Requirement
Need move Production Exchange setup along with DAG configuration to a new location.
IP Addresses for servers and DAG setup will change
No change in Server name and DNS names
During production server movement Mail services will be activated from DR Site
Need help in planning and executing the IP Address change/DAG Change
and Server movement.
Kumar K S

Hi,
I will have the following plans:
1. Active the DR site as the main site during the migration.
2. Moving the Server to the new site
3. Change the activate back to the Production Servers
https://social.technet.microsoft.com/Forums/exchange/en-US/fb9a27c3-81f8-4079-aeb8-42119b1e23bf/changing-ip-address-of-exchange-server
Thanks,
Simon Wu
TechNet Community Support

Exchange 2010 offline FSW failover occurs ?

Hello
We have two Exchange 2010 servers in DAG.
FSW is located on separate first DC and alternative FSW in located on second DC.
I know that Alternative FSW is not providing redundancy,...
We had a situation, that DC1 is been offline for few minutes because of windows updates
and when he was offline failover occurs from one Exchange node1 to node2 !
So my question is why failover occurs if the FWS is not online for some minutes and exchange 2 nodes are normaly online and working ?
Can we use Round Robin dns for FSW and create CNAME for fsw and add 2 hosts in it ? :)
Thanks for help,
BR
Marjan R.

Hello,
Sorry for delayed response.
You're right. The Alternate Witness Server itself does not provide any redundancy for the Witness Server, and DAGs do not dynamically switch witness servers, nor do they automatically start using the Alternate Witness Server in the event of a problem with
the Witness Server.
You can try to use Round Robin dns for FSW.
Cara Chen
TechNet Community Support

SAP Groupware Connector 5.0 Installation - Connect MS Exchange 2010

I try to install the SAP Groupware Connector 5.0 to connect a MS Exchange 2010 Server.
My host computer has an MS Windows Server 2008 operation system.
After starting the setup.exe and fill in the entries Exchange Server, Mailbox, User, Domain, Password I Get an error message:
"Validation of connection to Exchange Server failed"
"CDO library is not available."
"[0x80040154] Class not registered"
I am the opinion that I don't need a CDO, because I am useing a MS Exchange 2010 Server ....
Can somebody help me?!
Thanks

Hi Chrisitan,
Perhpas I wan'st clear in my previous reply. You cannot connect Groupware connector 5.0 to MS Exchange 2010 Server.
I've just checked, and found that exchange 2007 is now supported, but not 2010.Further information is available from the following link:
[http://help.sap.com/saphelp_crm70/helpdata/en/44/559f3699d908fae10000000a422035/frameset.htm]
Sorry I can't give a different answer....
Brian.

Exchange 2010 DAG Replication - too much data crossing the wire

I’m replicating 3 Exchange databases from our production active Exchange 2010 server across the WAN to another passive Exchange 2010 server at our DR site.
The Exchange server at the DR site does not have any active databases, i.e. no users are hitting that server.
We are running Update Rollup 8 for Exchange Server 2010 SP1 on both these servers.
The two sites are connected via a 10Mb/s MPLS connection and all the databases are in sync and the replication is working fine.
I have setup a network sniffer at the primary site and see a sizable amount of data crossing the wire from the production Exchange server to the DR Exchange server.
When I query the production Exchange server using the Tracking Log explorer and only choose the EventID ‘RECEIVED’ that should show me the amount of data that has been committed to the database.
If I choose a date range that is the same exact range that I have used to capture the raw data with my sniffer the amount of data the sniffer shows crossing the wire is 10 fold compared to what the Tracking Log Explorer shows.
If I actually count up the data in the LOG files it is about 20% more than what is crossing the wire but that seems to be because the DAG is compressing the data.
If in a one hour timeframe there is 500MB of data crossing the wire to the DR Exchange server the tracking log explorer will show only 50MB.
I would like to know why the data crossing the wire far exceeds the amount of data that is truly being sent/received from the primary exchange server.
Perhaps I’m simply not getting a true view of the amount of data being committed to the exchange server using the Tracking Log Explorer.
Maybe there is a better way to report how much data is being committed to the exchange databases.
Any assistance would be appreciated…

Two things.
1 - you must update Exchange to SP3 and a recent RU. Willard has already provided the links, which point back to my blog if you want to see the lifecycle map for Exchange 2010. SP1 has been out of support since January 2013. Time to move
on pllease
2 - Looking at the tracking log is not sufficient. I would not expect that to show everything.
I want to know what traffic you see as excessive. What ports are you seeing used here?
My money is on content indexing. CI will use additional traffic over and above log repl traffic. Expect CI traffic to be roughly the same again. You can test this by disabling CI on the database or stopping the services on the DR server.
To disable the CI for the database:
Set-MailboxDatabase DBName -indexEnabled $False
Or stop the Exchange search services on the DR box to leave production unaffected.
Again - you need to update Exchange. You would be better to do that now rather than when something breaks and Microsoft support cannot fully assist you since you are not uptodate.    I'll leave discussion of the security issues resolved
in recent Exchange RUs aside.
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.

Exchange 2010 WAN redundancy

Similar Messages

Maybe you are looking for