Trend and Alarm Backfilling

The trend and alarm backfilling use event queues that hold events until they are back filled into the Plant SCADA trend and alarm systems. The queue of events waiting to be processed for backfilling will be replicated between the drivers using the same TCP communications link that the cache replication uses.

The event queues in peer drivers are kept synchronized in the same manner as the driver cache is kept synchronized. That is, the image is copied on unit start-up, if there is a peer, and is kept synchronized by replication on queue state change.

There is no disk file image of the event queue. On cold start after a total system shutdown (shutdown of both peer drivers) the queue is empty.

Note: If some events have not been pushed to the alarm/trend servers and both I/O servers shutdown then the events and trend data will be lost.

Example

A device generates an event and sends it to the active I/O server. The active I/O server will then send the event to the peer I/O server. When it receives the event, the peer I/O server will add the event into its internal queue.

Then the active I/O server then adds the event to its internal queue before sending it to the alarm/trend server.

The alarm/trend server sends an acknowledgement to the active server on receiving the event. The active I/O server sends a message to the peer to remove the event from its internal queue. The event is then removed from the internal queue of the active I/O server.