Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase

AlwaysOn local HA with second site automatic failover. Expand / Collapse
Author
Message
Posted Thursday, April 3, 2014 12:54 PM
Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Yesterday @ 3:10 PM
Points: 306, Visits: 967
I'm looking for a solution to have cross data center automatic failover in the event of a data center loss for highly critical databases. I would like to have local HA and also automatic failover to the DR site. This does not seem possible with AlwaysOn.

Is my only option for automatic cross data center failover to build a node in one data center and a node in the other data center with a node/FS at a third data center in order to maintain quorum? I'd like to have local HA in the mix but that doesn't seem possible.

What pattern do you suggest for the highest data security and also availability?

Thanks
Post #1558188
Posted Thursday, April 3, 2014 4:53 PM
Ten Centuries

Ten CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen Centuries

Group: General Forum Members
Last Login: 2 days ago @ 9:57 AM
Points: 1,194, Visits: 2,234
When you say "cross data center automatic failover" are you refering to failing over whole instance\bunch of databases\a single DB ?

--
SQLBuddy
Post #1558309
Posted Thursday, April 3, 2014 6:15 PM
Old Hand

Old HandOld HandOld HandOld HandOld HandOld HandOld HandOld Hand

Group: General Forum Members
Last Login: Yesterday @ 3:10 PM
Points: 306, Visits: 967
I'm thinking of partially contained database so in the case the instance doesn't need to come across, just the database with the contained users.

Is it true that I would want one cluster node in one data center and the second cluster node in the other data center with a third node in a third data center in order to maintain quorum?
Post #1558330
Posted Thursday, April 3, 2014 6:41 PM
Ten Centuries

Ten CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen CenturiesTen Centuries

Group: General Forum Members
Last Login: 2 days ago @ 9:57 AM
Points: 1,194, Visits: 2,234
PHXHoward (4/3/2014)
I'm thinking of partially contained database so in the case the instance doesn't need to come across, just the database with the contained users.

Is it true that I would want one cluster node in one data center and the second cluster node in the other data center with a third node in a third data center in order to maintain quorum?


If it's a single database, then I don't think that's needed. Just use Failover Clustering in Data Center A and setup DB mirroring on that DB for the remote Data Center B.

--
SQLBuddy
Post #1558332
Posted Friday, April 4, 2014 5:28 AM


SSCertifiable

SSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiableSSCertifiable

Group: General Forum Members
Last Login: Today @ 9:53 AM
Points: 6,742, Visits: 14,381
PHXHoward (4/3/2014)
I'm looking for a solution to have cross data center automatic failover in the event of a data center loss for highly critical databases. I would like to have local HA and also automatic failover to the DR site. This does not seem possible with AlwaysOn.

To have local HA at the instance level would require a clustered instance, this may only participate in an AO group for manual failover. Given this, it doesn't meet your requirement.


PHXHoward (4/3/2014)
Is my only option for automatic cross data center failover to build a node in one data center and a node in the other data center with a node/FS at a third data center in order to maintain quorum? I'd like to have local HA in the mix but that doesn't seem possible.

What pattern do you suggest for the highest data security and also availability?

Thanks

You may still achieve local HA using alwaysOn by having a secondary replica on the primary site configured for automatic failover.


-----------------------------------------------------------------------------------------------------------

"Ya can't make an omelette without breaking just a few eggs"
Post #1558441
Posted Friday, April 4, 2014 5:53 AM


SSCrazy

SSCrazySSCrazySSCrazySSCrazySSCrazySSCrazySSCrazySSCrazy

Group: General Forum Members
Last Login: Thursday, December 4, 2014 10:09 AM
Points: 2,706, Visits: 3,427
First, make sure this database is not involved in any cross-database or distributed transactions. They are not supported with AO. Have you thought about doing a geo-cluster? We are currently doing this with one of our applications. It is a 3-node cluster with 1 node being in our DR site. For the storage, we use Site Recovery Manager (SRM).

Thanks,

Jared
SQL Know-It-All

How to post data/code on a forum to get the best help - Jeff Moden
Post #1558453
« Prev Topic | Next Topic »

Add to briefcase

Permissions Expand / Collapse