0

Maybe I'm searching for the wrong thing, but I can't see how I can do this.

I have two servers:

  1. Server A (has SQL Server 2012 installed)
  2. Server B (has SQL Server 2014 installed)

Server A provides an API via IIS. The code both inserts and gets data from the database depending on the API call.

Server B does exactly the same for its local database.

I would like to have it so that the two databases are consistent copies of each other. I do believe this is what HA accomplishes.

So a load balancer, for example, might direct API calls to A sometimes, and B sometimes, but the data needs to be consistent either way.

I found some documentation on HA for SQL Server 2012, but not on how to do this for a mixed environment of 2012 and 2014.

Any help or pointers to documentation would be great! Thanks.

migrated from stackoverflow.com Mar 18 '15 at 21:30

This question came from our site for professional and enthusiast programmers.

  • 2
    Are both 2012 and 2014 enterprise editions ? If yes, you can leverage AlwaysON - writes goes to primary and reads from readable secondary. Note that its not advisable to have HA between different versions (2012 and 2014) of sql server - because during disaster once you failover from lower (2012) to higher (2014) version, you cannot failback. – Kin Shah Mar 18 '15 at 20:46
1

how to do this for a mixed environment of 2012 and 2014.

Doing HA in a mixed environment is a recipe for disaster because during disaster once you failover from lower (2012) to higher (2014) version, you cannot failback.

Since you are using Enterprise edition of SQL Server, you can leverage AlwaysON technology wherein you can configure writes occurring on Primary and reads on the secondary configured as readable secondary.

Thinking a bit more on what you mean by mixed, I presume that transactional replication would be much better from sql server 2014 as publisher and 2012 acting as subscriber.

The performance would be dependent on how big is the database, if you are replicating only the tables that you need or the entire database, the latency between the two servers as well as the amount of transactions that are occurring on your publisher.

Note that if you go with T-Rep, doing schema changes will require a new snapshot. Also there will be a distribution database created as a part of setting up T-Rep.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.