Mirror for disaster recovery

  • Hi

    If I have SQL Server 2005 and want a up to date mirror of my database on another server, what's the best approach?

    If server 1 goes down I want to be able to switch to server 2. I don't require automatic failover. When server on goes up again I want automatic transfer of data from server 2 to server 1.

    Peer to peer replication?

  • Database mirroring?

    Log shipping?

    Clustering?

    Replication is not designed as a failover tool, it's more for availability.

  • Ah, you probably posted this after discovering mirroring is not yet supported. The short answer is to use log shipping.

  • Doesn't this mean that I have a time gap. Server 2 has the most resent backup of server 1?

  • Hmm.....Database Mirroring is already available. Earlier (still talking about 2k5 Beta) it was a notion that it is an option that could be taken from Microsoft, but it is already available in SQL Server 2005 (but i guess, ppl would still be skeptical until the Beta 3 release, maybe dec).  Mirroring = real time log shipping.

    I think a best disaster recovery would be - Clustering + Mirroring

    Anywyz, abt your question.....you will have the time gap which will be the time for bringing the mirror db online (unless u set up auto failover)

  • Database Mirroring is not to be used in production environment yet.

    Is log-shipping basically automatic restore of log-backups? Or have I misundertod?

    How often can I make log-backups without degrading performance horribly?

  • Don't forget to think outside the (SQL) box.

    Many sites use a disk replication product such as Double-Take or SAN replication to copy their SQL data to another site.  In this scenario the target site has Windows running but not SQL.  You use D-T to replicate the .mdf, etc files and any application files from source to target.  At failover time you stop SQL and D-T on the primary site and start SQL on your secondary site. 

    If you need to replicate a lot of flat file data as well as SQL databases, this approach can be better than SQL-based replication as it is easier to ensure that SQL and flat file data is kept in synch at the target.

    You should look at everything WHAT is needed to get your applications working at the target site before deciding on HOW you get the data there.

    Original author: https://github.com/SQL-FineBuild/Common/wiki/ 1-click install and best practice configuration of SQL Server 2019, 2017 2016, 2014, 2012, 2008 R2, 2008 and 2005.

    When I give food to the poor they call me a saint. When I ask why they are poor they call me a communist - Archbishop Hélder Câmara

  • Yes, and depends on how active the server is.

    First of all quantify your requirements (with the business), then pick a solution.

    How much downtime is acceptable?

    How much data loss is acceptable?

    Is manual intervention okay, and who will carry it out at 2a.m.?

    Does DR need to be geographically seperated (and by how far)?

    What are the cost implications of data loss and downtime?

    Etc.

Viewing 8 posts - 1 through 7 (of 7 total)

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