Always on with an SQL Cluster and 2 nodes

  • Hello all,

    I have searched everywhere and can't seem to find an example of setting up Always on where the primary server is an HA SQL cluster (2 nodes).

    I am in the process of setting up   Always on for the first time out side of  a lab.  Here is what my configuration looks like:

    Primary server =  SQL Clustered Server with 2 nodes (Active / Passive  Cluster)
    Secondaries =  1  Stand Alone DR server and 1 Stand Alone Reporting server.

    The request is to use Always on and replicate the DB's  on the SQL Cluster to both secondaries servers.  DR would be none readable (until an actual fail-over) and Reporting will be readable all the time.

    Is it possible to have always on where the primary server is a SQL Cluster server?
    If so, do I add the additional server to the existing Failover Cluster Manager?

    Does anyone have any links or how to's for this type of request?

    Thank you in advance.

    JZ

  • Joe Zonum - Friday, September 28, 2018 12:19 PM

    Hello all,

    I have searched everywhere and can't seem to find an example of setting up Always on where the primary server is an HA SQL cluster (2 nodes).

    I am in the process of setting up   Always on for the first time out side of  a lab.  Here is what my configuration looks like:

    Primary server =  SQL Clustered Server with 2 nodes (Active / Passive  Cluster)
    Secondaries =  1  Stand Alone DR server and 1 Stand Alone Reporting server.

    The request is to use Always on and replicate the DB's  on the SQL Cluster to both secondaries servers.  DR would be none readable (until an actual fail-over) and Reporting will be readable all the time.

    Is it possible to have always on where the primary server is a SQL Cluster server?
    If so, do I add the additional server to the existing Failover Cluster Manager?

    Does anyone have any links or how to's for this type of request?

    Thank you in advance.

    JZ

    Please check my stairway at this link

    http://www.sqlservercentral.com/stairway/112556/

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

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • Thank you very much.. I'll take a look.

  • Perry Whittle  -  Great write up....

    "Integrating technologies can have its advantages, but here, when using FCIs, there are clearly more disadvantages it would seem."

    Taking this to heart,  I would agree that the complexity my not be worth the pay off.  So I am thinking of dissolving SQL Cluster.

    Would you recommend evicting the secondary node from the cluster; so that is acts as a standalone server?
    Or would you recommend a rebuild.  The current server is not in production yet. So no users would be affected.

    JZ

  • Joe Zonum - Monday, October 1, 2018 11:15 AM

    Perry Whittle  -  Great write up....

    "Integrating technologies can have its advantages, but here, when using FCIs, there are clearly more disadvantages it would seem."

    Taking this to heart,  I would agree that the complexity my not be worth the pay off.  So I am thinking of dissolving SQL Cluster.

    Would you recommend evicting the secondary node from the cluster; so that is acts as a standalone server?
    Or would you recommend a rebuild.  The current server is not in production yet. So no users would be affected.

    JZ

    Are you planning to use an Availability Group in place of FCI?

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

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • Our new plan is to dissolve the Fail over cluster ( 2 node: Active\Passive).  Rebuild the server as a stand alone server with  additional nodes  DR\Reporting. Use Always on and AG readable replica's to DR and Reporting.

  • Joe Zonum - Thursday, October 4, 2018 10:18 AM

    Our new plan is to dissolve the Fail over cluster ( 2 node: Active\Passive).  Rebuild the server as a stand alone server with  additional nodes  DR\Reporting. Use Always on and AG readable replica's to DR and Reporting.

    ok, AGs require a windows server failover cluster unless you're planning a workgroup cluster or multi domain AG

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

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • Thank you  Perry, I am aware of the failover cluster requirement.I was able to setup a test lab and was able to successfully setup and test .

    The plan now is to log ship from Old Prod to New Prod, New Reporting, and New DR.
    The day of the cutover,  Recover the DB on New Prod and Setup Always on and the AG for the Prod DB to  New Reporting and New DR.
    I was able to test this in our lab and it worked as expected.

Viewing 9 posts - 1 through 8 (of 8 total)

You must be logged in to reply to this topic. Login to reply