3 Replies Latest reply on Jul 20, 2011 9:18 AM by ataylor

    HA without SAN?

    billy.sjoberg

      Hi guys,

       

      Might be a dumb question but note that I have read the manual and looked at (mostly all) forum posts regarding HA.

      We have an environment where we'd like to HA enable an AS6 with a backup server at a separate site. So we're only dealing with one live AS plus a passive backup AS + bundled Hornet instance on another node.

       

      I've been told by our infra-team that the two sites cannot share SAN's, the reasons for this evades me but that's supposedly just hard facts.

      Is there any way we can acheive HA failover using HQ 2.2.5 without resorting to a shared journal file system?

       

      Our requirement is the usual suspect; Having the backup server go active with the same persistent messages that existed on the live node after a failover in order to lose non-delivered messages.

        • 1. Re: HA without SAN?
          ataylor

          Replication is currently disabled in the current version, we are in the process of reenabling it and it will be in the next release

          1 of 1 people found this helpful
          • 2. Re: HA without SAN?
            billy.sjoberg

            That was a very quick answer Andy, thanks for the info!

             

            Ok, it starts to make sense to me now as I've seen discussions regarding the replication feature but it's not reflected in the latest manual.

            Reading an old manual (eg. http://hornetq.sourceforge.net/docs/hornetq-2.0.0.CR1/user-manual/en/html/ha.html) this makes sense.

             

            But this snippet is worrying me though:

            "One a live server has failed over onto a backup server, the old live server becomes invalid and cannot just be restarted. To resynchonize the pair as a working live backup pair again, both servers need to be stopped, the data copied from the live node to the backup node and restarted again.

            The next release of HornetQ will provide functionality for automatically synchronizing a new backup node to a live node without having to temporarily bring down the live node."

             

            Is the reason why you chose to remove it in order to rework it completely?

            If so, would the new functionality synchronize delivered messages while the primary was out as well in order to not duplicate messages?

            Say i.e. that

            - A had 50 undelivered persistent msgs in different queues during server crash

            - B becomes active and respective consumers picks up all messages

            - Failback occurs as A is fixed/restarted

            - A synchronizes deliveries performed by B, meaning that the messages won't be duplicated and redelivered?

            • 3. Re: HA without SAN?
              ataylor

              Is the reason why you chose to remove it in order to rework it completely?

              Yes amongst others, including dealing with split brain etc

               

              If so, would the new functionality synchronize delivered messages while the primary was out as well in order to not duplicate messages?

              Say i.e. that

              - A had 50 undelivered persistent msgs in different queues during server crash

              - B becomes active and respective consumers picks up all messages

              - Failback occurs as A is fixed/restarted

              - A synchronizes deliveries performed by B, meaning that the messages won't be duplicated and redelivered?

              what would happen is that on restart server A wouls start receiving replication packets from node B until it was synched, once this had happened failback would occur.