← Back to Articles
ServAssure NXT multidatacenter failover
Last Updated: June 30, 2026

This procedure outlines a failover procedure for the ServAssure NXT multidatacenter deployment scenario covered in this appendix.

Site Failover Site A to Site B

  1. Site A: Stop/disable the following services:
    • ServAssure Dashboard for Alarm Central (If installed)
    • Insight Reporting (If installed)
    • Trap Receiver
    • ServAssure Metrics Platform server
    • Tomcat
    • HA Proxy
    • ServAssure NXT engine
    • ServAssure Collection Manager
    • ServAssure Poller
    • RabbitMQ
    1. Run the playbook:
      ansible-playbook -i <inventoryFile-SiteA> tools/nxt-service-manager -u arris -bkK
    2. Cycle through each component and disable each service.
  2. Copy the ServAssure Collection Managerregistration.db to the collection managers at both Site A and Site B as a database "backup".
  3. Site B: Start/enable the following services:
    • HA Proxy
    • Solr
    • Zookeeper
    • RabbitMQ
    • ServAssure NXT engine
    • ServAssure Collection Manager
    • ServAssure Poller
    • Tomcat
    • Trap Receiver
    • Insight Reporting (Optional. The user web browser must be able to directly access this component if enabled.)
      NOTE: This component is optional. To support Insight Reporting access into the ServAssure NXT UI, ensure that users' web browsers can directly access Insight Reporting by enabling proper cross-origin resource sharing (CORS), secure HTTPS access, and configuring appropriate URL routing.
    • ServAssure Dashboard for Alarm Central
      NOTE: This component is optional. To support ServAssure Dashboard for Alarm Central access into the ServAssure NXT UI, ensure that users' web browsers can directly access ServAssure Dashboard for Alarm Central by enabling proper CORS, secure HTTPS access, and configuring appropriate URL routing.
    • ServAssure Metrics Platform server
    1. Run the playbook:
      ansible-playbook -i <inventoryFile-SiteB> tools/nxt-service-manager -u arris -bkK
    2. Cycle through each component and enable each service.
  4. Site B: Rebuild Solr index:
    1. Log in to the NXT-Engine administration UI.
    2. Select Administration.
    3. Click Solr Ingest for the target tenant(s).
  5. Site B: Start HA Proxy and reconfigure any load balancing rules if applicable.
  6. Site B: Run the site-upgrade.yml playbook from the ServAssure Collection Platform installation package. This will re-configure ServAssure Collection Platform which includes:
    • Configure ServAssure Collection Manager and ServAssure Poller hosts to connect to the RabbitMQ for Site B.
    • Ensure that the RabbitMQ, ServAssure Collection Manager, and ServAssure Poller systems are started.
  7. Re-enable any operational alerting for Site B.
    NOTE: The procedure for failing back over to the primary site would follow the same steps, only reversing Site A and Site B.