Fail Over to the Other Instance

You manually initiate a failover from the primary instance or from the secondary instance (if the primary instance is unreachable). You can perform failover during an actual disaster recovery occurrence or to test this functionality per your business or legal requirements.

Perform a Failover

Note:

No notification is sent by Oracle Integration when an instance fails. However, Oracle Cloud Infrastructure provides functionality for setting alarms when issues occur. See Managing Alarms.
  1. Go to the primary instance in the Oracle Cloud Infrastructure Console. For this example, failover is performed from the primary instance in one region to the secondary instance in a different region. The steps are the same when performing a failover from the secondary instance.
  2. Click Start failover.


    The Integration instance details page is shown. A green circle labeled OIC is shown as Active on the left. The instance name has a label of Primary. Below this are tabs for Service console, Edit, Failover, Move, and Add tags, and a More Actions drop-down list. Below this, the tab for Integration instance information is shown. Labels are shown for Created, Updated, Version, Consumption model, Edition, OCID, Service console URL, Shape, License type, Message packs, File Server, and Disaster recovery.

  3. Click Failover when prompted to fail over to the secondary instance in the other region.

    The process to fail over to the instance in the other region begins. A second tab opens in your browser for failover status about the secondary instance.

    Because data synchronization between the two instances has occurred in near real time since the completion of disaster recovery installation, the failover process takes approximately the same amount of time regardless of the amount of design-time metadata data in your instance.

  4. Follow failover progress in the Work Requests section.

  5. When failover completes, the previous primary instance goes into standby mode. The word Secondary now appears below the instance name. The status changes to STANDBY below the circle labeled OIC.

  6. Click the tab in your browser to access the new primary instance. The status is shown as ACTIVE below the circle labeled OIC and the word Primary now appears below the instance name. The word _Recovery remains appended to the end of the new primary name.

    If you were logged in to the primary instance of Oracle Integration during failover, you receive the following message indicating that the primary instance is now in a standby state.


    The message "This is a standby instance" is shown. Below the message says "Please go to primary to access console." A URL link to the new primary instance is provided.

  7. Click the global (regionless) URL specified in the message to log in to the new primary instance. The global URL does not mention any region name.
  8. Continue working as you were prior to receiving this message. Activated integrations in the original primary instance are displayed as activated in the new primary instance. As you work in the new primary instance, your changes to data are synchronized with the original primary instance automatically.
  9. When the original primary instance is restored, click Failover in the Oracle Cloud Infrastructure Console of either instance if you want to fail back to the original primary instance.

Deactivate and Reactivate Integrations with Polling Triggers

When you initiate a failover for integration instances running with polling endpoints (for example, integrations involving Oracle Cloud Infrastructure Streaming, databases, JMS, and others), the integration transfers to the failover instance but is not automatically activated. After failover, you must manually identify and reactivate these integrations. Once reactivated, message flow resumes as expected.

Change the Built-in API Calls to Reflect the New Hostname and Integration Instance Name

If you use the Oracle Integration built-in APIs, the hostname and integration instance name change in the API call after failover completes.

For example:

  • Pre-failover API call: 
    https://mydesign-pp-integration-prod-gen3.integration.us-ashburn-1.ocp.oraclecloud.com/ic/api/integration/v1/integrations/
?integrationInstance=sysqa-drtest-cgs-inst2-bxffubagcv29-to-pp
  • Post-failover API call: 
    https://mydesign-pp-integration-prod-gen3.integration.us-phoenix-1.ocp.oraclecloud.com/ic/api/integration/v1/integrations/
?integrationInstance=sysqa-drtest-cgs-inst2-remote-bxffubagcv29-yu-pp

This is expected behavior because the global (regionless) URL is used only for runtime. You must change the design-time hostname and integration instance name in the API call to reflect the post-failover values.