FSW and DAG/DAC questions

Z

[Zid]

#1


Hi, we are having two datacernters (different buildings) that we want to place our new Exchange 2010 environment in.

Datacenter #1
1 x HUB/CAS with a File Share Witness
1 x Mailbox as a DAG member holding 2 Active Databases and two Passive Databases)

Datacenter #2
1 x HUB/CAS with an Alternate File Share Witness
1 x Mailbox as a DAG member holding 2 Active Databases and two Passive Databases)

To use the DAC feature I would have to have 3 or more Mailboxservers to be members of a DAG?

If Datacenter #1 failes, how can I make the alternate fileshare witness to be active/primary?

If MBX1 fails the MBX2 has contact with FSW in DC1 and will mount the rest of the databases that were mounted on the failed server.
If DC1 looses connection with DC2, and we want to run on the services provided in DC2, is there a manuall way to force FSW and the services in DC2?
 
M

Michel de Rooij

#2


With RTM that was a requirement, but with SP1 DAC also works for 2 DAG members. The alternate FSW can be preconfigured so that it can be picked up during a datacenter *over (more info via same link).

More info here (part of my 3 part DAC blog):
http://eightwone.com/2010/09/01/changes-in-exchange-2010-sp1/

More info on datacenter switch/failover and related procedures:
http://technet.microsoft.com/en-us/library/dd351049.aspx

Michel de Rooij,
MCITP Ent.Msg 2007+2010| MCTS W2008, Ex2007+2010 Conf, OCS2007 Conf | MCSE+Msg2k3 | MCSE+Inet2k3 | Prince2 Fnd | ITIL
I blog on http://eightwone.wordpress.com/ and tweet on http://twitter.com/mderooij
 
Z

[Zid]

#3


Hi Michel,

In my schenario, would you recommend using/enable DAC?

Trying to read your blog and technetarticle and am I getting this right:

With alternate FSW we dont have to specity the Set-DatabaseAvailabilityGroup <DAGName> -WitnessServer <ServerName> during the activation of the database servers since it has been taken care of using the alternateFSW feature?

We only have to:

1. terminate the cluster and evict the member in the failed datacenter

2. restart the cluster service in the second datacenter and use /forcequorum to start the service

3. evict the failed node member

4. Restore-DatabaseAvailabilityGroup will use the alternate FSW instead of former one specified? (configure FSW is not needed here if we are in DAC mode?)
 
Z

[Zid]

#4


Hi Michel,

In my schenario, would you recommend using/enable DAC? We will defenately go for SP1!

Trying to read your blog and technetarticle and am I getting this right:

With alternate FSW we dont have to specify the Set-DatabaseAvailabilityGroup <DAGName> -WitnessServer <ServerName> during the activation of the database servers since it has been taken care of using the alternateFSW feature?

We only have to:

1. terminate the cluster and evict the member in the failed datacenter

2. restart the cluster service in the second datacenter and use /forcequorum to start the service

3. evict the failed node member

4. Restore-DatabaseAvailabilityGroup will use the alternate FSW instead of former one specified? (configure FSW is not needed here if we are in DAC mode?)
 
M

Michel de Rooij

#5


I suggest using DAC when possible, less chances of split brain and simplified procedures for switch/failover & recovery.

The alternative FSW is picked up when predefined so you don't need to specify it when using Restore-DatabaseAvailabilityGroup.

Michel de Rooij,
MCITP Ent.Msg 2007+2010| MCTS W2008, Ex2007+2010 Conf, OCS2007 Conf | MCSE+Msg2k3 | MCSE+Inet2k3 | Prince2 Fnd | ITIL
I blog on http://eightwone.wordpress.com/ and tweet on http://twitter.com/mderooij
 
Z

[Zid]

#6


Followup question: is this accurate even if we are using both datacenters as active? We have active databases on both locations to split the load!
 
Top