Using the DNP3 Slave Driver on a Redundant Slave

[Anonymous user]

📖 Home  Back  
The filtered Value Maps which may be attached to the slave outstation will store events, up to the configured limit (by default 50), for the attached points when they cross thresholds etc. These events are stored in the data stream of the database and are transferred to any standbys. Each event has two flags associated with it.

The first flag indicates that the event has been read by the DNP3 Driver process. This flag is NOT transferred to any standbys and is not saved. The second indicates that the event has been processed and acknowledged by the remote DNP3 master. This flag is transferred to any standby servers. The first flag prevents the event being transferred to the DNP3 Driver process on this main for a second time. However if a server change over (or server restart) occurs, this flag will not play any part in assessing which events need to be transferred to the new DNP3 driver process. Only the second flag will be used to consider which events need to be transferred to the new DNP3 driver process (i.e. all unacknowledged events).

If the Value Map exceeds its configured limit for events, it will first remove the oldest acknowledged events followed by the oldest events read by the driver, before finally removing the oldest events that have not been read by the driver if there is still not enough space. If this is occurring frequently (indicated by a buffer overflow event at the DNP3 master end) you should consider increasing the maximum number of events the Value Map can store.

Caution should be observed when restarting the DNP3 driver process without restarting the server as events read by the previous driver instance but not acknowledged will not be made available for the new driver instance to send.

Go: Home Back