This documentation applies to the 8.1 version of BMC Atrium Core, which is in "End of Version Support." You will not be able to leave comments.

To view the latest version, select the version from the Product version menu.

Configuring Atrium Integrator Engine in failover mode

The following topics are provided:

Note

Failover is supported only for the Atrium Integrator engine. Atrium Integrator engine is deployed with BMC Remedy AR system server installation. The Atrium Integrator component that is installed later, deploys the plugin-ins required to transfer data is BMC Atrium CMDB.

Failover considerations

Consider the following when you configure Failover mechanism for Atrium Integrator engine:

  • Failover is supported only for the primary server.
    If the primary server is up and if the secondary server goes down, all the jobs and transformations running on the secondary server will not be failed-over. For example, if for manual load-balancing purpose you specify Job A to run on the Atrium Integrator engine 2 (secondary server) and if Atrium Integrator engine 2 goes down while Atrium Integrator engine 1 (primary server) is up, Job A will not be failed over. 
  • Suspended mode is not supported for Atrium Integrator engine. At one time, engines on all the nodes will be running.

  • This capabilty is available only for version 8.1.02 and later. 

Configuring Atrium Integrator engine in failover mode

The following diagram illustrates how Atrium Integrator engineis configured in a failover environment. 

To configure Atrium Integrator engine in a failover mode

  1. When you install BMC Remedy AR System server on all the nodes of the server group, Atrium Integrator engine is also installed on each node. 
  2. Open the AR System Server Group Operation Ranking form. 
  3. Update the Atrium Integrator engine ranking. See, Setting failover rankings for servers and operations

    Note

    Ranking of the Atrium Integrator engine and the BMC Remedy AR Server can be different.  

  4. Open the UDM:Config form. See, Configuration form
  5. Verify that the Atrium Integrator Engine field, displays the AR server connect name for all the servers in the server group.
  6. Restart the BMC Remedy AR system server.
  7. Open the UDM:Config form and verify if the Is Default check box is selected for the primary Atrium Integrator engine server.

Job and transformation failover process

To manage the job and transformation failover process, the Carte servers log entry of all the new and in progress jobs and transformations in the UDM:CartePending form. The secondary Atrium Integrator engineuses this information to execute new and in progress jobs and transformations of the primary Atrium Integrator engine that went down. For more information about UDM:CartePending form, see Carte pending form

The following diagram illustrates the job and transformation failover process followed with the explanation: 

  1. Atrium Integrator engine 1 (primary server), Atrium Integrator engine 2 (secondary server), and Atrium Integrator engine 3 (tertiary), log the entries of new and in progress jobs and transformation into to UDM:CartePending form. 
  2. When Atrium Integrator engine 1 goes down, Atrium Integrator engine2 is treated as the primary server. Accordingly, The Is Default check box is automatically selected for Atrium Integrator engine 2 in the UDM:Config form. The ranking happens according to order specified in the AR System Server Group Operation Ranking form.
  3. Atrium Integrator engine 2, refers to the UDM:CartePending form to see if there were any new or in progress jobs or transformations on Atrium Integrator engine 1 when it went down. 
  4. Atrium Integrator engine 2 picks up the new or in progress jobs or transformations from Atrium Integrator engine 1 and re-runs them. Here:
    • Atrium Integrator engine 2 will run the job and transformations from the beginning and not from where Atrium Integrator engine 1 went down.
      For example, if a job was scheduled to run for 6 hours on Atrium Integrator engine 1, and if Atrium Integrator engine 1 went down after 3 hours, Atrium Integrator engine 2 will run the job all over again from the beginning and will take as long as it normally would - in this case six hours from restart on the second Atrium Integrator engine.
    • However, in case of delta jobs, Atrium Integrator engine 2 will run the job from the point Atrium Integrator engine 1 went down.
    • Atrium Integrator engine 2 will not copy the cache from Atrium Integrator engine 1. Fresh cache will be generated on Atrium Integrator engine 2. 
    • While the Atrium Integrator engine 2 re-runs the job, no duplicates are created thus, there is no impact to the integrity of data in BMC Atrium CMDB. 
    • Also, while Atrium Integrator engine 2 is running the failed over job or transformation, if Atrium Integrator engine 1 comes up, the job or transformation will be still completed on Atrium Integrator engine 2. However any new requests will be served by the rank 1 server thereafter. 
    • If Atrium Integrator engine 2 server goes down after Atrium Integrator engine1 is up and running, the in-progress job or transformation that was picked up from Atrium Integrator engine 1 server will not again fail over.

Note

To run scheduled jobs you must select the default Carte server (object NGIECarteserver) when adding a schedule. See, Managing Atrium Integrator job schedule.

Troubleshooting

To troubleshoot Atrium Integrator engine failover issues, refer to following log files:

  • For server failover communication issues, refer to the <AR installation directory>\BMC Software\ARSystem\ServerGroupLogForm file.
  • For post failover operation issues, refer to the <AR installation directory>\BMC Software\ARSystem\Carte-signal.log file.

This version of the documentation is no longer supported. However, the documentation is available for your convenience. You will not be able to leave comments.

Comments

  1. Ramesh Anbuthambi

    When the Primary server come up, why the switch over is not happening to Primary Server? I see in UDM:Config form, the secondary server still set as Default Carte server.

    Feb 20, 2018 08:31