4 Replies Latest reply on Sep 5, 2017 5:58 AM by jameseydoyle

    Protection Group consideration

    ok7827125 Lurker

      I have installed SRM 6.1 with EMC RecoverPoint SRA on two different sites. in this scenario I have installed PSC, VC, and SRM machines separately in  two sites and our PSCs are configured in Linked Mode.

      One of my doubts is the correct location for PSC, VC and SRM virtual machines. In fact, I want to know whether these virtual machines would be in the Protection LUN or not?

      I currently put these machines in the same place of other protected virtual machines. is it correct or not?

      the second point I want to add is that after failing the primary site and  running a recovery plan, we successfully fail over in the second site. after that, we have problem to fail back. Should the LUN be presented on the first site automatically by SRM or manually through RecoverPoint?

       

       

       

        • 1. Re: Protection Group consideration
          p_hall Hot Shot
          VMware Employees

          SRM/VC/PSC VMs should be separated from the protected VMs. These VMs should not be failed over during disaster recovery or planned migration.

           

          Regarding your second point - assuming your storage has been configured correctly in the first place, SRM should then handle everything during failover/failback. i.e. storage that was unpresented or made read-only during the failover, will be automatically re-presented or made read-write during failback.

           

          Were you able to reprotect successfully after failing over?

          • 2. Re: Protection Group consideration
            jameseydoyle Enthusiast
            VMware Employees

            Hi,

             

            You should not include your vCenter in a protection group as all your failed over VMs will be added to the remote site vCenter's inventory anyway. So after failover, your production site vCenter will have nothing left to manage. Also, you want to keep your production vCenter in situ so that you have a valid target for the reprotect and failback.

             

            For the reprotect operation to function successfully, a completed planned failover needs to be performed. If your actual DR event forced you to use the 'Disaster Recovery' option rather than the 'Planned Migration' option, there may have been errors encountered that were ignored to allow the failover to complete. By performing a 'Planned Migration' in the same direction, any tasks that completed successfully will be skipped, but any tasks that encountered errors will be re-run until they are successful. Then you can be sure that your original site in a prepared state for the reprotect. If this option is not possible, it may be necessary to manually reverse the replication of the LUN through RecoverPoint.

            • 3. Re: Protection Group consideration
              ok7827125 Lurker

              I powered off all Hosts all of a sudden in the main site and after that I have run Recovery plan in Disaster recovery mode. fortunately, all VMs powered on in the disaster site. after a while, the main site will be fixed and online and now I 'm gonna to return to the main site. unfortunately, reprotect button is not enable and it is grayed. Although I had to reverse it from the Recoverpoint and got the Image Access in the first site, I did not get any result and I can' t fail back. the status in recovery plan shown Recovery is required !

              do you have the same issue?

               

              • 4. Re: Protection Group consideration
                jameseydoyle Enthusiast
                VMware Employees

                Hi,

                 

                I replied to your previous post before seeing this additional post. If you ran in full 'Disaster Recovery' mode (i.e. Forced Recovery mode as exposed by the advanced setting), then only Recovery site operations are performed and no attempt is made to communicate with the original site, so SRM databases are likely out of sync with one another. When the original site somes back online, the prepareFailover SRA command should be issued for any production devices that are still reporting as Source LUNs to make them read-only. You could attempt a 'Planned Migration' but you may never get it to complete, due to the nature of the 'Forced Recovery'. If this is the case, you will need to manually reverse the replication and recreate a new Recovery Plan using the same devices as before.