Always On with Cluster + Standalone Question/Discussion

  • In the next couple of months we are finally getting a much needed upgrade to SQL from 2005 to 2012. I'm really looking forward to some of the new features particularly 'Always On Groups'. My company is sending me on a course in a couple of weeks to go over said features but in the mean time I was doing some digging.

    The plan is to migrate our current Clusters over but I also wanted to leverage AO for reporting purposes. So...if I read the documents correctly it appears like I should be able to create an AO group between the clusters and a standalone instance (provided they are connected to the same WSFC) correct?

    Anybody here with experience have any tips or gotcha's regarding this? Any input is appreciated.

    Thanks,


    SELECT quote FROM brain WHERE original = 1
    0 rows returned

  • As far as I know, this is correct. You can have some AG nodes in an FCI and some not. I didn't see an MSDN reference, but probably one somewhere.

    I did find this: http://www.mssqltips.com/sqlservertip/3150/adding-sql-server-alwayson-availability-groups-to-existing-failover-clusters/

  • That's a real good article. It does seem like I'm on the right track but sometimes it's the little caveats in the wording of some MSDN that get me.

    e.g. https://msdn.microsoft.com/en-us/library/ff929171.aspx

    More specifically:

    An availability replica can be hosted by either a standalone instance of SQL Server or an FCI instance. Only one FCI partner can host a replica for a given availability group.

    I would assume that in a failover the new active node would become the new host of the replica and not break the Availability Group. Unless I missed it, it's just not that clear. In the end I'm sure Microsoft will be able to address my questions during my classes but it's always good to do a little homework. 😉


    SELECT quote FROM brain WHERE original = 1
    0 rows returned

  • Really looks more like this to be honest

    Microsoft do detail FCIs within AlwaysOn groups, check the following link

    https://msdn.microsoft.com/en-us/library/ff878487.aspx#FciArLimitations

    Also check my article at this link

    http://www.sqlservercentral.com/articles/FCI/92196/[/url]

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

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

  • yb751 (4/2/2015)


    I would assume that in a failover the new active node would become the new host of the replica and not break the Availability Group.

    Correct. There would of course be an outage while the failover of the FCI occurs, depends on how long the FCI failover takes.

    There is a default timeout on the AlwaysOn group but since failover from a replica that is an FCI is a manual process the Primary database could potentially be offline for sometime, especially if the FCI does not come online for whatever reason.

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

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

  • Perry Whittle (4/3/2015)


    yb751 (4/2/2015)


    I would assume that in a failover the new active node would become the new host of the replica and not break the Availability Group.

    Correct. There would of course be an outage while the failover of the FCI occurs, depends on how long the FCI failover takes.

    There is a default timeout on the AlwaysOn group but since failover from a replica that is an FCI is a manual process the Primary database could potentially be offline for sometime, especially if the FCI does not come online for whatever reason.

    Thanks for the response Perry and great write up on FCI and AO. Could you just clarify what you meant regarding the manual failover? In my example I plan the FCI to hold down the same role it always has. Data feeds and mission critical applications will still hit the FCI regardless of what node is holding the resources. The other standalone instance would be a read only replica for reporting.

    I'm going to make some (probable) assumptions again and go with the FCI instance will still function and failover the same as it always has otherwise what would be the point. Do you mean that JUST the AO replica will need to be manually failed over? So for example there is X problem and a failover occurs on the FCI. Everybody is still happy since the instance is still accessible. Meanwhile those who are hitting the reporting node are still fine but...it is no longer being updated until the host replica is manually failed over?

    Sorry...just want to know what I'm in for.

    Thanks for the info!


    SELECT quote FROM brain WHERE original = 1
    0 rows returned

  • Is DTC a requirement?

  • JoshDBGuy (4/6/2015)


    Is DTC a requirement?

    Off the top of my head I'd say no...long answer is I'm not 100% sure. I'm still discovering many 'interesting' things left/done/created by my predecessors.


    SELECT quote FROM brain WHERE original = 1
    0 rows returned

  • yb751 (4/6/2015)


    Do you mean that JUST the AO replica will need to be manually failed over?

    Yes exactly that. The FCI will behave as usual.

    If the FCI is unavailable then failover within the AO group is a manual process

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

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

  • Awesome, thank you sir!


    SELECT quote FROM brain WHERE original = 1
    0 rows returned

Viewing 10 posts - 1 through 9 (of 9 total)

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