Configure failover for the connector

High availability for the connector can be achieved by configuring two instances of a connector as a failover pair. Each instance communicates its own failover state and its data source status to the other instance over the administrative port. Currently, failover for HART-IP can be configured only in "warm" mode, which requires that each connector can communicate with the data source simultaneously.

To configure failover, each instance of the connector (the active instance and the backup instance) should be configured to write data to the same PI and PI AF servers. Also, in order to write data into the same PI points and PI AF elements the data source name and configuration must be the same for both connector instances.

1. Open the Failover pane of the connector administration page.

2. Click the Enable failover hyperlink to enable failover for the connector.

3. Enter the Peer connector Host name or IP address and Peer Connector Port (this is the administration port).

   The peer host is the instance of the connector that will be paired with the current connector.

   Note: Make sure the current connector and peer are configured so that both use host names or both use IP addresses to identify each other but not a mixture of the two. Mixing host name and IP address could result in miscommunication between the connectors and lead to data loss.

4. Repeat the steps 1 through 3 by logging into the peer connector administration page.

   After the connector is started, each connector communicates with its own peer.

Failover can also be enabled after the connector is started; however, to avoid sending duplicate data to Data Archive and PI AF servers, you must configure the first connector with peer information and then start the other connector in the failover pair.