Redundant Mode

To reinforce the s ecurity of the whole system, it is possible to use two computers in redundancy.

PC n°1 PC n°2 SCADA ALERT ALERT SCADA PC's network PLC's

To perform a full redundancy , t he Alert software and the supervision software (or the communication server with the PLC network) must be installed on each of the two computers, with the same configuration. The two computers supervises a unique database (the PLC network can be doubled) and are interconnected by a computer network. Each of the Alert application supervises its own environment. As the PLC data base is the same, each computer detects the same alarms. In redundant mode, only the active computer transmits calls. Acknowledgements performed on a computer are automatically transmitted to the other one. The determination of the active compute r is automatic. The two computers supervise themselves reciprocally through the computer network. A computer considers itself in active state if there is no fault detected (communication with the supervisor OK and at least a communication system in service). If the two computers are active at the same time, the computer not declared as main station changes to the inactive state. The secondary station becomes active only when the main station becomes inactive or if it can not supervise this one normally (main station or network out of order). It is possible to define some supervision tags as specific to one station (declared as "no redundant" tags). This allows the definition of events as specific to an environment or issuing from the other station. An alarm declared as no redundant triggers the call independently of the alert state (active or inactive) of the station.

The updating of the configuration of the two computers can be automatically synchronized: schedule tables, supervision list (redundant variab les), messages, and alarm and call parameters. This synchronization can be configured. To insure an optimal resumption when switching from a station the other, alarms, acknowledgments and call cycles can also be synchronized. It is also possible to perfor m a reduced redundancy by installing only the Alert software on the secondary computer. This last will have for function to trigger an alarm call if the main computer goes in defect. The main station also can supervise the secondary station of the same way. To trigger a call in case of defect of the main station (station out of order or incapable to call), it is enough to supervise the "REDUNDANCYFAIL" item of the DDE server of Alert, by declaring an alarm on this item (with the event condition "different of 0"). The redundancy can be reinforced by the declaration of redundant data servers. For each data server used (DDE or OPC), it is possible to define one or more redundant servers. In case of failure of a data server, the software automatically restore s the connections with the next data server in the list of redundant servers of the failed server.