SQL Server Developer Center > SQL Server Forums > SQL Server Disaster Recovery and Availability > How do I change a passive node to an active node?
Ask a questionAsk a question
 

AnswerHow do I change a passive node to an active node?

  • Thursday, November 05, 2009 4:21 PMCaptSky Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     

    Current Situation

    SQLCluster1                           SQLCluster2     NOTE: Both C1 and C2 are separate servers

    SQLVCluster1                        SQLvCluster2

    Active

     

    Passive

     Current Resource Groups

    Q:\  Quorum

    R:\  Database Data --- In use

    S:\  Database Logs --- In use

    T:\  WebCT Data ---------- Available for use

    U:\  WebCT Logs ---------- Available for use

     

    Where We Must Be

    SQLvCluster1                        SQLvCluster2

    Active

     

    Active

     

    We have no money for new Severs, SANs, etc., until July 1st 2010.

    So, I must turn SQLvCluster2 into an Active node – I have no other choice! 

     

    I would greatly appreciate the help of anyone who can assist me.

     

    Thank you,

     

    CaptSky

Answers

  • Thursday, November 05, 2009 9:21 PMTome Tanasovski Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     Answer
    Once you add the resources to the new group you can rt-click on the group and select move group to move those resources to the other node..... 

    For example:

    Cluster Group 1:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    Quorum
    Cluster Name
    Cluster IP

    Cluster Group 2:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    R-drive
    S-drive
    VIP
    Name
    SQL Service (MSSQL Server)
    Agent
    etc

    Cluster Group 3:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    T-drive
    U-drive
    VIP
    Name
    SQL Service (WEBCT Instance)
    Agent
    etc

    Now you can make sure that cluster group 1 and 2 run on the node SQLCLuster1 while cluster group 3 will run on SQLCuster2.  This is an active-active cluster that will use the horsepower of both servers.  In the case of a hardware failure the groups that are on the failed node will fail over to the other node.... allowing both instances to run on a single node during the emergency....  You may get a performance hit at this time, but you will not be completely out.


    • Marked As Answer byCaptSky Friday, November 06, 2009 2:04 PM
    •  

All Replies

  • Thursday, November 05, 2009 4:33 PMTome Tanasovski Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     
    Create a new cluster group and add the resources you want to run on the other node.  Each cluster group can then be failed over independantly.
  • Thursday, November 05, 2009 8:45 PMCaptSky Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     
    Ok, so just by adding a new cluster group is going to turn my Passive node into an Active Node?  I don't understand how that will happen.
  • Thursday, November 05, 2009 9:21 PMTome Tanasovski Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     Answer
    Once you add the resources to the new group you can rt-click on the group and select move group to move those resources to the other node..... 

    For example:

    Cluster Group 1:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    Quorum
    Cluster Name
    Cluster IP

    Cluster Group 2:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    R-drive
    S-drive
    VIP
    Name
    SQL Service (MSSQL Server)
    Agent
    etc

    Cluster Group 3:
    Possible Owners: SQLCluster1, SQLCluster2
    Resources:
    T-drive
    U-drive
    VIP
    Name
    SQL Service (WEBCT Instance)
    Agent
    etc

    Now you can make sure that cluster group 1 and 2 run on the node SQLCLuster1 while cluster group 3 will run on SQLCuster2.  This is an active-active cluster that will use the horsepower of both servers.  In the case of a hardware failure the groups that are on the failed node will fail over to the other node.... allowing both instances to run on a single node during the emergency....  You may get a performance hit at this time, but you will not be completely out.


    • Marked As Answer byCaptSky Friday, November 06, 2009 2:04 PM
    •  
  • Friday, November 06, 2009 2:04 PMCaptSky Users MedalsUsers MedalsUsers MedalsUsers MedalsUsers Medals
     
    Ok, that makes much more sence,

    Thank you,

    CaptSky