Watchdog Groups

Watchdog groups are meant for connection monitoring. Each watchdog group contains only one item and expects that the item sends values at regular intervals n. If the item does not transmit a value within 3*n milliseconds, the watchdog group sets the allItemsInvalid bit.

After a watchdog is triggered, the driver does not reset it after the first received value. The driver waits until values are received for the defined time validToReset without any interruption.

Configuration

Use the [opc_<symbolic_servername>] watchdogGroup config entry to configure the watchdog group, the interval n and validToReset. To define the interval n, a multiplication factor must be given. The interval n is calculated via the given factor and the actual update rate of the OPC group (in milliseconds). The update rate of the OPC group is written on the UpdateRateAct data point element of the respective _OPCGroup data point.

The watchdog mechanism can also be used for items which do not transmit values unsolicited. E.g. if a script starts a refresh or read operation for the item in the group every n-2 seconds and the group records the received values. If there are not enough values received, allItemsInvalid is set and thus the communication error is detected. (Attention: reading processes must be performed to DEVICE or this mechanism will not work, since there will be values from cache available all the time.)

Example

[opc_server1]

watchdogGroup = "ABB_Watch” 2 10

With the defined multiplication factor and an actual update rate of 1000 milliseconds, the interval n would be 2000 milliseconds. If e.g. a counter is used in the periphery, it must be set to an increment time that is less than 2000 milliseconds.

In case that the watchdog is triggered, the clients must receive values for at least 10 seconds until the watchdog is reset. E.g. if values are received for only 8 seconds, an internal counter is reset and values must be received for 10 seconds again. This prevents the watchdog from being continuously set and reset.

Restrictions

  • If the update rate granted by the server is lesser than the watchdog interval, the group will signal this by setting the allItemsInvalid state.

  • Watchdog groups are always permanent groups. Being identified in the config file as a watchdog group is stronger than being a nonpermanent group.

  • Watchdog groups will only accept one item since they monitor one connection.

  • It is not possible to create a new watchdog group at runtime, the client must be restarted.

  • In case of a redundant system the actual value of the item has no meaning since it can be overwritten by the active system on the passive system.