exchange 2013 DAG | Standalone database

Posted on 2014-11-10
Last Modified: 2014-11-24

In my test lab I have 4 exchange servers

Svr1 - Site 1
Svr2       - Site 1
Svr3       - Site 2
Svr4       - Site 2

All of the above servers are configured in a DAG.

Datacentre activation co-ordination is enabled.

Svr 4 has a database called DB1 that is not part of the DAG and has some test user accounts in it.
I kill the WAN link and go to my Windows 8.1 test vm that is also in site 2.  I log on using one of the accounts that has a mailbox in DB1 and open outlook.  The autodiscover process works but after that it fails to connect to DB1.

I also expected applications to be able to relay through either svr4 or svr3 whilst the WAN link is down but this also doesn’t seem to work.  I have an application receive connector on all the exchange servers.
Question by:cmatchett
  • 3
  • 3
  • 2
  • +1
LVL 41

Expert Comment

ID: 40432759
You are combing two different concept into one. Think again, what are you trying to achieve.
LVL 19

Expert Comment

by:Adam Farage
ID: 40432763
The autodiscover process works but after that it fails to connect to DB1.

So two things.. 1) is the database actually online (Get-MailboxDatabase DB1 -Status) and 2) what is the AutoDiscoverServiceInternalUri set to (Get-ClientAccessServer | Select Identity, AutoDiscoverServiceInternalUri)

Most cases when a full site fails you need to change some DNS A records, which is usually along with If your autodiscover namespace internally is the same all around, you might think AutoDiscover is working but its not. You have two options:

1) Setup a separate namespace in each site for AutoDiscover (so the primary site is and the other active site is or
2) Manually change the A record for Autodiscover to point to the surviving sites CAS or CAS Virtual IP (for load balancing)

As for the send / receive connectors I would need more information on how that is actually configured. In theory you should have a send connector to the internet or smarthost (which then sends to the internet) that has all four CAS within there. If that does fail then you are set, as the other two surviving CAS would still utilize it.
LVL 37

Accepted Solution

Jamie McKillop earned 166 total points
ID: 40432765

If you have a DAG that spans sites, the DAG can only be active in one site. If you lose your WAN link, quorum will be lost in one of the sites and the databases will dismount. You say DB1 is not part of the DAG but if it resides on a DAG member it is part of the DAG.

If you want both sites to remain active when the WAN link fails, you need to create a separate DAG for each site.


Author Comment

ID: 40432786
my fault on the applications relay, i hadn't checked anonymous.

If i want to use site 2 for a DR test and the DR test involved killing the WAN link, i could remove svr4 from the DAG and add it again once the WAN link is restored?  DB1 would then mount on SVR4 whilst the WAN link is down?
Best Practices: Disaster Recovery Testing

Besides backup, any IT division should have a disaster recovery plan. You will find a few tips below relating to the development of such a plan and to what issues one should pay special attention in the course of backup planning.

LVL 19

Expert Comment

by:Adam Farage
ID: 40432792
is the database down when you kill the WAN link? I would imagine so since the cluster manager is going to take all the databases offline due to lack of quorum.
LVL 41

Assisted Solution

Amit earned 167 total points
ID: 40432810
As I said above, you need to revisit your concept. You are trying to mix DAG with Standalone server, however in your case, it is just standalone DB running on a DAG. So, if your DAG failover or down, it will impact the standalone as well, as it is running on DAG server, Until you remove server from DAG mix, it will behave same for standalone database.

Author Comment

ID: 40432854
OK.  I should be able to mount the database under the condition that i remove the server from the dag?
LVL 19

Assisted Solution

by:Adam Farage
Adam Farage earned 167 total points
ID: 40432888
Yes. the reason it will not mount when the two other nodes are not available is because quorum is not met, which will block the Store.exe service from allowing the databases to start up.

If you remove the mailbox server as a node from the DAG then this would resolve the issue, but I am not 100% sure why you would have a DAG and not want to have a mailbox database with no copies within it.

Author Closing Comment

ID: 40463150
great help

Featured Post

Make managing Office 365 email signatures a breeze

Are you using Office 365? Having trouble trying to set up email signatures for your users? Getting stressed out managing multiple signatures? Need an easier way to manage? We have a solution for you, try the most-user friendly and powerful signature management tool on the market.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Follow this checklist to learn more about the 15 things you should never include in an email signature from personal quotes, animated gifs and out-of-date marketing content.
Find out what you should include to make the best professional email signature for your organization.
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
This video discusses moving either the default database or any database to a new volume.

919 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

16 Experts available now in Live!

Get 1:1 Help Now