UNIFYBroker Disaster Recovery and Cold Standby

In order to properly plan for disaster recovery scenarios, the UNIFYBroker state may be fully recovered in a timely manner. The main components requiring backup are the configuration files in the Extensibilitydirectory. Through these, the UNIFYBroker state can be restored by adding these files to a service instance, and running full connector imports to restore the contents of the entity repositories. Note that this will result in new entity identifiers being created for each user - this should be kept in mind when this field is being used for join criteria or other role in the solution.

As an additional precautionary step, namely in cases where the entity identifier is of importance or the datasets in target systems are significant in size, the database may be backed up and restored.

UNIFYBroker can be configured to operate in cold standby by configuring an additional instance with the same configuration files, keeping in mind the resolution of DNS names, SQL aliases and IP addresses from the standby machine. The Data Configuration should also point to the same database used by the live instance. In the event that the live instance goes offline, the second instance can be enabled and will continue functioning as expected. Any and all connections to the old UNIFYBroker instance will need to be updated to reference the new service.

ALERT: Services in cold standby should not be enabled, or should have all connector and adapter operations disabled, otherwise contention issues may occur.

Is this article helpful for you?