The following forum(s) are migrating to a new home on Microsoft Q&A (Preview): Azure Virtual Machines!

Ask new questions on Microsoft Q&A (Preview).
Interact with existing posts until December 13, 2019, after which content will be closed to all new and existing posts.

Learn More

 none
SQL Server HADR in Azure RRS feed

  • Question

  •  

    I'm trying to understand what we'd get out 'out of the box' for a SQL Server 2012 install on an Azure VM (Iaas) to determine if we need to do more (ie availabiltiy groups)

    You apparently get 3 copies of your VM automatically.  Under what conditions would you want to use an Availability Group with synchronous replication and automatic failover instead?

    • Is VM failure automatically detected?
    • What happens is MS performs maintenance?
    • What do these 3 copies do for me?

    Geo-replication provides and additonal 3 copies of your vm in a remote data center. Under what conditions would I need to use an AG with asynchronous replication instead?  I understand that one case is if the data base files + log span more than one disk.

    Tuesday, September 23, 2014 10:16 PM

Answers

  • Hi,

    Thanks for your post.

    For the SQL Server HADR in Azure, please refer to the article below:

    High Availability and Disaster Recovery for SQL Server in Azure Virtual Machines

    http://msdn.microsoft.com/en-us/library/azure/jj870962.aspx

    Regards.


    Vivian Wang

    Wednesday, September 24, 2014 6:03 AM
  • Is VM failure automatically detected? Yes, but there is not failure detection inside the VM. If SQL Server fails, it will not be detected.

    What happens if MS performs maintenance? the VM becomes unavailable for 12-15 minutes.

    What do these 3 copies do for me? It is not 3 copies of the VM, it is 3 copies of the storage. This guarantees no data loss.

    when to use Availability Group: Within same data center, can use to detect errors and for fast failover. AG failover takes 15 seconds, as opposed to the 12-15 minutes for VM failover. Across different regions, the AG provides fast failover as opposed to up to 24 hours for Azure disaster recovery. With AGs, you can test failover of events to ensure that it properly works.

    Wednesday, September 24, 2014 5:23 PM
    Moderator

All replies

  • Hi,

    Thanks for your post.

    For the SQL Server HADR in Azure, please refer to the article below:

    High Availability and Disaster Recovery for SQL Server in Azure Virtual Machines

    http://msdn.microsoft.com/en-us/library/azure/jj870962.aspx

    Regards.


    Vivian Wang

    Wednesday, September 24, 2014 6:03 AM
  • Is VM failure automatically detected? Yes, but there is not failure detection inside the VM. If SQL Server fails, it will not be detected.

    What happens if MS performs maintenance? the VM becomes unavailable for 12-15 minutes.

    What do these 3 copies do for me? It is not 3 copies of the VM, it is 3 copies of the storage. This guarantees no data loss.

    when to use Availability Group: Within same data center, can use to detect errors and for fast failover. AG failover takes 15 seconds, as opposed to the 12-15 minutes for VM failover. Across different regions, the AG provides fast failover as opposed to up to 24 hours for Azure disaster recovery. With AGs, you can test failover of events to ensure that it properly works.

    Wednesday, September 24, 2014 5:23 PM
    Moderator
  •   Thanks.  some follow questions onthe 24 hour DR turnaround

      At a very high level how to you recover from a DC outage with Geo-replication? Do you need to rebuild a VM and attach data and then re-attach the data files ? Does MS do this for you?

     Does Geo-replication just replicate the data disk or does it also replicate the C: drive?

    thanks

    Peter

    Wednesday, September 24, 2014 5:40 PM