Fail over group on Azure SQL without down time. RRS feed

  • Question

  • Hi Guys,

    We have 2 Server;

    Server A  And Server B

    Server A is Primary and Server is Secondary and in read only Mode and part of Failovergroup as set.

    Scenario 1:

    When we are deploying we want to deploy only on secondary server disconnecting failure, so that we dont have any down time on the primary server A

    For that we will be removing the Server B which is in secondary sate from failure over group and deploying code.

    Once my deployment is complete on Server B, can SERVER A and SERVER B should be synced without data loss?



    Navin.D http://dnavin.wordpress.com

    Wednesday, June 19, 2019 8:01 PM

All replies

  • Hi Navind,

    In a planned or manual fail-over scenario, the following occurs:

    •  Planned Failover

    Planned failover performs full synchronization between primary and secondary databases before the secondary switches to the primary role. This guarantees no data loss. Planned failover is used in the following scenarios:

    1. Perform disaster recovery (DR) drills in production when the data loss is not acceptable.
    2. Relocate the databases to a different region.
    3. Return the databases to the primary region after the outage has been mitigated (failback).

    • Manual failover

    You can initiate failover manually at any time regardless of the automatic failover configuration. If automatic failover policy is not configured, manual failover is required to recover databases in the failover group to the secondary. You can initiate forced or friendly failover (with full data synchronization). The latter could be used to relocate the primary to the secondary region. When failover is completed, the DNS records are automatically updated to ensure connectivity to the new primary.

    Auto-failover group terminology and capabilities (link). 

    Please let us know if you have additional questions.



    Thursday, June 20, 2019 4:44 PM
  • Hi Navind,

    If I got your point... This is the sequence:

    1) Break Auto failover group (where is your application pointing? to the listener?)

    2) Deploy your code (or implementantion) on the server B

    3) if everything is fine, you want to recreate the auto failover group, right? because if you are trying to do this you will always loss data... Since direct seeding will start again and the database cannot exist on the new secondary replica (you will have to drop it first)... Unless you have a process to identify which tables changed and then move those records to the other part you will loss data.

    Hope this helps!


    Monday, November 11, 2019 1:07 PM