Results 1 to 5 of 5
  1. #1
    Join Date
    Feb 2009
    Posts
    24

    Unanswered: Manual Scripts to Perform HADR Failover

    I need to setup automating failover for live HADR clusters.
    Environment: Windows 2008 R2 64 bit, DB2 v9.7 ESE FP 3a
    It seems I cannot use a solution like Tivoli System Automation for Multiplatforms. (license issue)
    I might need to implement manual scripts for doing this task asap ...
    I searched through the internet and I did not find anything suitable ...
    Is this possible to create .sql scripts or can you do this using Task Center by creating different tasks that perform some checks?
    The big problem is what to check in order to decide when to perform the failover?
    1. To check if you can connect to DB's
    2. To connect and also to list applications to see if there are any applications running also?
    (it is possible that applications run but you cannot connect ... different scenarios ... )
    Also I am concerned in setting up for HADR: SYNC, NEARSYNC, ASYNC and HADR_PEER_WINDOW value...
    (these are servers sitting in the same rack - recommended SYNC, two Ethernet connections)

  2. #2
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Are you sure TSA MP is not licensed for db2 on windows? As far as I gather, TSA is licensed for a two node cluster when you have High Availability Feature Pack. Without HA Feature Pack, you cannot do HADR.

    On Windows, the TSA installation is not integrated with the db2 installation, but provided as a separate install.

    I could be wrong though
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  3. #3
    Join Date
    Aug 2001
    Location
    UK
    Posts
    4,650
    Coming to your original question, of course you can write your own scripts to monitor HADR and perform takeovers when required. In effect, you are developing a cluster software-both monitoring part and the agent. So, do not underestimate the complexity and effort required to get a proper set of scripts working.
    Your monitoring part is more complex than the agent part. If you have another cluster software that provides a facility to develop your own agents, then you can of course do it.

    a) Availability of Server
    b) Availability of Instance
    c) Network interconnect between the pair.

    These are the three areas that hadr monitoring focusses on.
    Based on your monitoring, you have to decide when each of these incidents have occurred, and take appropriate action.

    Take into account the 'split-brain' scenario.

    HTH atleast to some extent

    Sathyaram
    Visit the new-look IDUG Website , register to gain access to the excellent content.

  4. #4
    Join Date
    Feb 2009
    Posts
    24
    Thank you Sathyaram for your advices!
    Yes I would like to see how someone did such scripts, make the checks and take the decision to takeover standby database making sure that no split-brain scenario would occur.
    If someone can post such scripts here I would really appreciate.

    Thank you,
    Regards,

  5. #5
    Join Date
    Jan 2009
    Location
    United Kingdom
    Posts
    77
    sathyaram , I agree there is a lot of complexity. Ihave tried in the past , but found TSA very useful , for the heartbeat moniotoring and automatic failover.
    heartwork - you can see the HADR scipts ion the TDA folders , assuming the installation has come with that feature - if not you should be able to download

Posting Permissions

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