Post reply

Disaster recovery in SQL 2005 - which one is better ?

  • January 27, 2010 at 5:52 am

    #137614

    Hi All,

    i am in a slight confusion of implementing a DR solution for one of my banking client. They have a citrix application which depends on a sql server. they cannot have a down time more than 15 mins for that application...They want the DR database to come up in less than 15 mins and i am in a slight confusion which DR solution to go with...

    I had ruled out Clustering as the client is not willing to bear the cost of it. I also ruled out database mirroring as we failed to get the required performance benchmark with synchronous mirroring with RTT of 167 ms

    I am in confusion whether to go with log shipping or replication. if i am going with replication which should i go for.

    Could anybody please help me to get a better solution ?

    My environment is Windows 2003+ SQL 2005 Standard edition

    Regards

    Nimesh

  • January 27, 2010 at 6:20 am

    #1109227

    Both of them can be used for your DR, but what matters in replication is, it does not take care of replicating the entire database as and when you make changes (add new Tables,Stored procedures, views etc). You will need to have an additional publication for the new Database objects..

    With your case, my choice would be to think on implementing Log Shipping..

    Do some reading and follow other forum members advice.

    Bru Medishetty

    Blog -- LearnSQLWithBru

    Join on Facebook Page Facebook.comLearnSQLWithBru

    Twitter -- BruMedishetty

  • January 27, 2010 at 6:40 am

    #1109243

    what about asynchronous mirroring? Perhaps they should improve their network.

    I would never use replication for DR, thats not really what it is intended for.

    which leaves log shipping or replication at the drive level (SAN mirroring)

    If these are the backend SQL databases that support citrix, log shipping would be fine

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

  • January 27, 2010 at 8:42 am

    #1109350

    One issue with replication is, if you're using transactional replication and the downtime is caused by someone running

    DELETE

    FROM ReallyCriticalTable

    -- Oops, forgot the where clause

    this is going to take down the replicated system as well.

    With log shippping you can use point in time restore to just before this point

Viewing 4 posts - 1 through 3 (of 3 total)

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