locked
Failover clustering RRS feed

  • Question

  • Hi all,

    how to automatically connect client  to passive node when active node fail over happened?

    thanks

    Wednesday, June 12, 2013 8:43 AM

Answers

  • Hi,

    http://www.brentozar.com/archive/2012/02/introduction-sql-server-clusters/

    When you connect to a cluster, you do so via a virtual name.  This name resolves to the currently active node.  When a failover happens, that name resolution changes to the new active node, but the name itself doesn't change.  So, your clients don't have to do anything other than refresh their connection. 

    Did you have something more specific in mind?



    Thanks, Andrew

    • Proposed as answer by Fanny Liu Thursday, June 13, 2013 7:47 AM
    • Marked as answer by Fanny Liu Thursday, June 20, 2013 8:15 AM
    Wednesday, June 12, 2013 9:06 AM
  • I would like to say some extra word on what Bainbridge already explained.

    When failover happend, DNS entry (which is a mapping between Virtual Cluster name and IP address of Active node) will be updated with new active node's IP address. So that when client connectes via the virtual name, name resolve will happen to point new IP address. No changes are necessary in application level to handle/response to failover.


    Srinivasan

    • Proposed as answer by Fanny Liu Thursday, June 13, 2013 7:47 AM
    • Marked as answer by Fanny Liu Thursday, June 20, 2013 8:15 AM
    Wednesday, June 12, 2013 9:44 AM

All replies

  • You should have connected to the virtual server name ( defined during the setup) and hence it is not a problem at all

    Best Regards,Uri Dimant SQL Server MVP, http://sqlblog.com/blogs/uri_dimant/

    MS SQL optimization: MS SQL Development and Optimization
    MS SQL Blog: Large scale of database and data cleansing
    Remote DBA Services: Improves MS SQL Database Performance

    Wednesday, June 12, 2013 9:00 AM
  • Hi,

    http://www.brentozar.com/archive/2012/02/introduction-sql-server-clusters/

    When you connect to a cluster, you do so via a virtual name.  This name resolves to the currently active node.  When a failover happens, that name resolution changes to the new active node, but the name itself doesn't change.  So, your clients don't have to do anything other than refresh their connection. 

    Did you have something more specific in mind?



    Thanks, Andrew

    • Proposed as answer by Fanny Liu Thursday, June 13, 2013 7:47 AM
    • Marked as answer by Fanny Liu Thursday, June 20, 2013 8:15 AM
    Wednesday, June 12, 2013 9:06 AM
  • I would like to say some extra word on what Bainbridge already explained.

    When failover happend, DNS entry (which is a mapping between Virtual Cluster name and IP address of Active node) will be updated with new active node's IP address. So that when client connectes via the virtual name, name resolve will happen to point new IP address. No changes are necessary in application level to handle/response to failover.


    Srinivasan

    • Proposed as answer by Fanny Liu Thursday, June 13, 2013 7:47 AM
    • Marked as answer by Fanny Liu Thursday, June 20, 2013 8:15 AM
    Wednesday, June 12, 2013 9:44 AM