Failover clustered instance with availbility groups

  • This question is just a sanity check

    approach 1:

    Prod - shared storage between server 1 and 2

    Server1: clustered SQL instance with availability group as primary

    Server2: Passive server for clustered instance of PROD

    DR - shared storage between server 1 and 2

    Server1: Clustered SQL instance with availability group as replica

    Server2: Passive server for clustered instance of DR

    Approach 2: Using replicated SAN

    Prod -

    Server 1: Standalone instance with availability group as Primary

    Server 2:Standalone instance with availability group as replica

    DR -

    Server 1: Offline until Disk group 1 (Prod server 1) has been broken and brought online at DR

    Server 2: Offline until Disk group 2 (Prod server 2) has been broken and brought online at DR

    Both these approaches will work wont they? I have only built and played with normal availability groups across servers, not mixing it with clustered instance \ replicated SAN

  • SQLAssAS (10/22/2014)


    This question is just a sanity check

    approach 1:

    Prod - shared storage between server 1 and 2

    Server1: clustered SQL instance with availability group as primary

    Server2: Passive server for clustered instance of PROD

    DR - shared storage between server 1 and 2

    Server1: Clustered SQL instance with availability group as replica

    Server2: Passive server for clustered instance of DR

    This details an FCI in the prod and an FCI in the DR with an Availability group between them, correct?

    If this is the case remember that you won't be able to configure automatic failover between the replicas. Why on earth do you want to have a DR instance that's clustered? DR is meant to be a short term usage until your live site is back online, totally unnecessary IMHO.

    Also by using FCIs you're just re introducing the storage single point of failure

    SQLAssAS (10/22/2014)


    Approach 2: Using replicated SAN

    Prod -

    Server 1: Standalone instance with availability group as Primary

    Server 2:Standalone instance with availability group as replica

    DR -

    Server 1: Offline until Disk group 1 (Prod server 1) has been broken and brought online at DR

    Server 2: Offline until Disk group 2 (Prod server 2) has been broken and brought online at DR

    Both these approaches will work wont they? I have only built and played with normal availability groups across servers, not mixing it with clustered instance \ replicated SAN

    Would need to know a lot more about your storage configuration, but why are you constantly pulling in shared or replicated storage to the design? AlwaysOn does not promote or indeed require replicated or shared storage. Also by using replicated storage you're just re introducing the storage single point of failure.

    The data replication is done at the software level. All nodes\replicas require simple, local, fast disk storage, spend time and money on boosting your network connectivity.

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

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

  • Perry Whittle (10/22/2014)


    SQLAssAS (10/22/2014)


    This question is just a sanity check

    This details an FCI in the prod and an FCI in the DR with an Availability group between them, correct?

    If this is the case remember that you won't be able to configure automatic failover between the replicas. Why on earth do you want to have a DR instance that's clustered? DR is meant to be a short term usage until your live site is back online, totally unnecessary IMHO.

    Also by using FCIs you're just re introducing the storage single point of failure

    Yes automatic failover to DR will always a manual step if we have 2 servers at PROD for resilience so this is not too much of an issue for us. But thanks for clarifying that. Single point of failure on the storage is something i did not think about though! I will look closer at our storage set up to see what reslience we have here.

    Would need to know a lot more about your storage configuration, but why are you constantly pulling in shared or replicated storage to the design? AlwaysOn does not promote or indeed require replicated or shared storage. Also by using replicated storage you're just re introducing the storage single point of failure.

    The data replication is done at the software level. All nodes\replicas require simple, local, fast disk storage, spend time and money on boosting your network connectivity.

    If we use a cluster on Prod it will obviously give server resilience and the availability group to DR gives automatic DR soltution if implemented properly. Thats the only reason I wanted use a cluster. I take your point on the Cluster at PROD and Cluster at DR as being overkill. THis is just due to the nature of the service it will be running. We would like ot have PROD and DR like for Like.

    Thanks alot for your feedback it has given me a few things to think about.

  • PROD Server 1 – Fixed storage, primary for High availability group, Automatic failover

    PROD Server 2 – Fixed storage, secondary for High availability group, read only copy

    DR Server 1 – Fixed storage, high availability group replica, read only copy

    That is clearly the traditional use of an availbility group for High availability / DR.

    What methods do people use to keep MSDB and master in sync accross the 3 servers? So jobs and permissions?

  • SQLAssAS (10/28/2014)


    PROD Server 1 – Fixed storage, primary for High availability group, Automatic failover

    PROD Server 2 – Fixed storage, secondary for High availability group, read only copy

    DR Server 1 – Fixed storage, high availability group replica, read only copy

    That is clearly the traditional use of an availbility group for High availability / DR.

    That represents a standard AlwaysOn group configuration, no shared storage and automatic failover available.

    SQLAssAS (10/28/2014)


    What methods do people use to keep MSDB and master in sync accross the 3 servers? So jobs and permissions?

    Most common method would be via scripts.

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

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

Viewing 5 posts - 1 through 4 (of 4 total)

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