4 Replies Latest reply: Mar 6, 2013 11:56 PM by CKPT RSS

    dataguard swithover_status "failed destination" issue

    995353
      oracle version: 11Gr2
      I have a RAC which has two nodes(node1, node2) as primary and a single instance(named DR) as standby. I didn't realize that node2 is down when I did a switchover test. Now the swithover_status of node1 is "to primary", the switchover_status of DR is "failed destination".

      There are some logs in alert.log:TNS-12543:TNS:destination host unreachable. caused by node2 is down.

      My question is how can I fallback to RAC as primary without node2 running? What configuration need to do?

      Edited by: 992350 on Mar 6, 2013 9:04 PM

      Edited by: 992350 on Mar 6, 2013 9:09 PM

      Edited by: 992350 on Mar 6, 2013 9:09 PM
        • 1. Re: dataguard swithover issue
          CKPT
          992350 wrote:
          I have a RAC which has two nodes(node1, node2) as primary and a single instance(named DR) as standby. I didn't realize that node2 is down when I did a switchover test. Now the swithover_status of node1 is "to primary", the switchover_status of DR is "failed destination".
          What is the version?
          When you perform switchover, the only one node of primary and one node of standby should up, remaining will be closed[down]
          There are some logs in alert.log:TNS-12543:TNS:destination host unreachable. caused by node2 is down.
          So can you try to ping the node2 host? are you able to ping or not?
          If this node completely decommissioned, then you can set CLUSTER_DATABASE to FLASE and then you can proceed.
          My question is how can I fallback to RAC as primary without node2 running? What configuration need to do?
          First check the switchover_status and database_role from v$database from primary and standby database.
          • 2. Re: dataguard swithover_status "failed destination" issue
            CKPT
            .... Duplicate post ....

            Edited by: CKPT on Mar 7, 2013 10:48 AM
            • 3. Re: dataguard swithover issue
              995353
              When I perform switchover, only node1 and DR is running.
              It's unable to ping node2 now. But node2 will be recovered in future.

              The swithover_status, database_role of DR is "failed destination", "primary" now.
              The swithover_status, database_role of RAC is "to primary", "physical standby"
              • 4. Re: dataguard swithover issue
                CKPT
                When I perform switchover, only node1 and DR is running.
                It's unable to ping node2 now. But node2 will be recovered in future.
                If right now not in use, then go ahead and disable cluster database as "CLUSTER_DATABASE=FALSE"
                The swithover_status, database_role of DR is "failed destination", "primary" now.
                The swithover_status, database_role of RAC is "to primary", "physical standby"
                Where is output? If you give just word its not easy to determine, Moreover after you perform switchover now both primary and standby are in SYNC?
                Complete the synchronization, disable cluster database and then you can perform switchover.

                You can refer MAA & manual documents
                http://www.oracle.com/technetwork/database/features/availability/maa-wp-10gr2-switchoverfailoverbest-128455.pdf
                http://docs.oracle.com/cd/B19306_01/server.102/b14230/sofo.htm