2 Replies Latest reply on Jan 29, 2020 7:48 PM by Paul Robins

    Control-M: On premise high availability design and effort estimation on a high level

    Satya Chatterjee
      Share This:

      Dear SMEs,

      Greetings!

       

      Just looking for a high availability design of new implementation of Control-M on premise and effort estimations with step wise at a high level on

       

      - Planning

      - Design

      - Test

      - Implement

      - Jobs migration as is

       

      If anyone of us has worked on the above, a document/ deck would be helpful masking the confidential information.

      Thanks in advance.

        • 1. Re: Control-M: On premise high availability design and effort estimation on a high level
          Keyur Patel

          Hi Satya,

           

          Control-M 9.0.x has inbuilt high availability, I am not sure which version you are working on, but if it is Control-M 9.0.x then best option to use Control-M high availability option.

           

          High level Desgin plan,

          • Separate servers for Control-M EM and Control-M server (Primary & Secondary)
          • Separate server for database
          • Install Control-M EM & Control-M server as primary on one server (by pointing to separate database server)
          • Install as Secondary Control-M EM & Control-M server on 2nd server (by pointing to separate database server)

          High lever Test plan,

          Control-m HA has zero down time, so you can plan your testing while jobs are running to fail over to secondary instance and fail back to primary server ( Shutting down primary EM\server will fail to secondary instance, by turining on Primary will take over from secondary instance, while jobs are running)

           

          Note: you have to check firewall rules to allow port communication between this serves, you can refer to BMC documentation for port details

           

          Check BMC Documentation for more details: Control-M 9.0.18.200 Control-M Documentation - Control-M - BMC Documentation

           

          Hope this will be helpful to you..

          Thanks

          Keyur

          3 of 3 people found this helpful
          • 2. Re: Control-M: On premise high availability design and effort estimation on a high level
            Paul Robins

            Note that 'zero downtime' is a little misleading for Control-M Server.

            When an agent starts receiving job requests from a different authorised Server, by default it takes 15 minutes before it starts responding. So jobs that are in progress when CTM Server failover occurs from primary to secondary could remain in progress for up to 15 minutes longer than normal.

            Regards,

            Paul.

            2 of 2 people found this helpful