Transactional Replication

  • You also have define Disaster. It can be considered a disaster if a harddrive fails, or switch dies, a server crashes. A disaster does not have to mean the destruction of the server room, or the building.

  • opc.three (6/7/2012)


    jitendra.padhiyar (6/7/2012)


    OPC.Three:

    You are correct that clustering can be used. And I said clustering will ensure hardware availability, but its not 100% DR, thats why I said clustering alone can not be said DR plan.

    Correct me if I am wrong.

    Thanks,

    Jitendra

    Why are you hung up on 100%? Do you think Log Shipping or Mirroring are a 100% DR solution?

    I am thinking in the same line as Jitendra, if we configure clustering in the same location without log shippping/mirroring, how can we say that clustering is a DR solution. It just means that if something goes wrong with one node it will automatically failover. Also, another disadvantage with clustering is that you only have one copy of data, where as with mirroing/log shipping you can have multiple copy of data. I am not trying to make an argument here, just trying to get all the facts correct....Please correct me if I am wrong.....with anything I mentioned above. I am always here to learn.

    Regards,

    TA

    Regards,
    SQLisAwe5oMe.

  • There is also the other part of Disaster Recovery, how much is the company willing to spend, how much are they willing to lose (data wise), how quickly do they need to be back up (your SLA).

  • SQLCrazyCertified (6/7/2012)


    opc.three (6/7/2012)


    jitendra.padhiyar (6/7/2012)


    OPC.Three:

    You are correct that clustering can be used. And I said clustering will ensure hardware availability, but its not 100% DR, thats why I said clustering alone can not be said DR plan.

    Correct me if I am wrong.

    Thanks,

    Jitendra

    Why are you hung up on 100%? Do you think Log Shipping or Mirroring are a 100% DR solution?

    I am thinking in the same line as Jitendra, if we configure clustering in the same location without log shippping/mirroring, how can we say that clustering is a DR solution. It just means that if something goes wrong with one node it will automatically failover. Also, another disadvantage with clustering is that you only have one copy of data, where as with mirroing/log shipping you can have multiple copy of data. I am not trying to make an argument here, just trying to get all the facts correct....Please correct me if I am wrong.....with anything I mentioned above. I am always here to learn.

    Regards,

    TA

    Lynn is making a point on this topic that hopefully will resonate with you. The something you mention can be defined as a disaster. It depends on what that something is, i.e. what you are trying to protect from as to which technology will be relevant. Something does not always mean total site-loss. All three technologies can play a part.

    There are no special teachers of virtue, because virtue is taught by the whole community.
    --Plato

  • Lynn Pettis (6/7/2012)


    There is also the other part of Disaster Recovery, how much is the company willing to spend, how much are they willing to lose (data wise), how quickly do they need to be back up (your SLA).

    Completly agree with you Lynn, that it is upto SLA of company that will choose the DR plan and how much they are ablt to spend on that. If company able to spend money and want good DR plan than clustering with Mirroring/Logshipping and use of correct version of RAID are the best practice to protect data and have high availabiltiy as per my knowledge.

    Thanks,

    Jitendra

  • opc.three (6/7/2012)


    SQLCrazyCertified (6/7/2012)


    opc.three (6/7/2012)


    jitendra.padhiyar (6/7/2012)


    OPC.Three:

    You are correct that clustering can be used. And I said clustering will ensure hardware availability, but its not 100% DR, thats why I said clustering alone can not be said DR plan.

    Correct me if I am wrong.

    Thanks,

    Jitendra

    Why are you hung up on 100%? Do you think Log Shipping or Mirroring are a 100% DR solution?

    I am thinking in the same line as Jitendra, if we configure clustering in the same location without log shippping/mirroring, how can we say that clustering is a DR solution. It just means that if something goes wrong with one node it will automatically failover. Also, another disadvantage with clustering is that you only have one copy of data, where as with mirroing/log shipping you can have multiple copy of data. I am not trying to make an argument here, just trying to get all the facts correct....Please correct me if I am wrong.....with anything I mentioned above. I am always here to learn.

    Regards,

    TA

    Lynn is making a point on this topic that hopefully will resonate with you. The something you mention can be defined as a disaster. It depends on what that something is, i.e. what you are trying to protect from as to which technology will be relevant. Something does not always mean total site-loss. All three technologies can play a part.

    Thanks for all your inputs guys.

    TA.

    Regards,
    SQLisAwe5oMe.

  • You're welcome.

    There are no special teachers of virtue, because virtue is taught by the whole community.
    --Plato

Viewing 7 posts - 16 through 21 (of 21 total)

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