Results 1 to 4 of 4
  1. #1
    Join Date
    Nov 2007
    Posts
    72

    Unanswered: Db2 High Availability

    HI ALL,

    O/S- AIX 5.3 WITH RAID 5
    DB VERSION- DB2 ENTERPRISE SERVER EDITION 8.2.3
    OTHER TOOLS - DATA PARTITIONING FEATURE ,SAN BOX for data storage.


    We will hdve clustering(ACTIVE-ACTIVE) among 4 machines and HACMP also.

    Q1.In such a scenario is it required to use DB2 HADR or REPLICATION techniques ? i am not sure but i feel its not required ?
    HACMP will it self be sufficient for high availability ?
    Q2.What should be the backup strategy since we are in OLAP environment ?
    we take full offline backup on fridays and incremental delta on weekdays and we have log retain and userexit on .is it correct to have userexit on when we are taking an offline backup ?
    Q3. During the restoration of the database the new target database should have userexit turned on ?
    Q4.does LOGARCHMETHOD1 does the same as userexit ? then why userexit is there ? i had compiled userexit2 using a c compiler ? what is the difference ?

  2. #2
    Join Date
    May 2003
    Location
    USA
    Posts
    5,737
    Q1. HADR provides recovery from a complete failure of the primary database. That includes the server and disk subsystems. HACMP uses shared disk, so if the disk subsystem goes out (or you have a double disk failure on a RAID array) then you could have a problem. HADR has a complete duplicate of the server and all the data (you need to make sure the standby is on a completely different disk subsystem than the primary for complete protection).

    Q2. You can have log retain on for an offline backup. This will allow you to roll the logs foward after a restore (but you need some way to copy or give accesss to the logs to the target server).

    Q3. That depends on whether you need to use the rollforward command.

    Q4. LOGARCHMETH1 basically does the same thing as the userexit sample program. Userexit is still provided for compatibility reasons and for those whose exit program is more sophisticated than simply moving the archive logs to another path.
    M. A. Feldman
    IBM Certified DBA on DB2 for Linux, UNIX, and Windows
    IBM Certified DBA on DB2 for z/OS and OS/390

  3. #3
    Join Date
    Nov 2007
    Posts
    72
    but i am using san box for my storage and only softwares will be installed on the participating clustering machines ? is it the correct way i am thinking ? why backup the san box again and where to back up ?

  4. #4
    Join Date
    May 2003
    Location
    USA
    Posts
    5,737
    HADR is like Oracle's Data Guard, in that it provides complete redundancy of the server and the data. Do whatever you want.
    M. A. Feldman
    IBM Certified DBA on DB2 for Linux, UNIX, and Windows
    IBM Certified DBA on DB2 for z/OS and OS/390

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •