Understand Flatline Watchdog

The Flatline Watchdog application monitors all Flatline Watchdog (watchdog) records to determine if the telemetered value trend of any configured monitored point has matched or satisfied a specified trigger condition.

The following monitored points are permitted:

• Analog point

• Status point

When the application detects that the trigger condition has been matched, the watchdog record is triggered. The points associated with the watchdog record are marked as stale and a watchdog alarm is generated.

The Return-to-Normal (RTN) state requires the following to occur:

• The Flatline Watchdog application must evaluate to good condition for the number of application processing cycles (not necessarily consecutive) during the configured time interval. These processing cycles are configured through RTN Hold Off Cycles (where the default is 0) and RTN Hold Off Interval (default is zero). Once the Flatline Watchdog returns to a normal state, the points associated with the watchdog record are evaluated for a fresh condition on the next poll cycle or a scheduled integrity update, and then the watchdog alarm is cleared.

When the application detects that the trigger condition of a monitored value has been reached, it evaluates the RTN Hold Off Cycles and RTN Hold Off Interval values. If those values are not defined, then the trigger condition is cleared, the points associated with the watchdog record become fresh on the next poll cycle or scheduled integrity update, and the alarm is cleared.

When RTN Hold Off Cycles and RTN Hold Off Interval are defined, then the RTN Hold Off Cycles number of polls from the monitored point has to match the not-to-be triggered values before the RTN Hold Off Interval expires to clear the trigger condition, put the associated points in fresh, and clear the alarm.

If you select Ignore RTN Hold Off on a monitored point, the application ignores the RTN Hold Off Cycles and RTN Hold Off Interval settings. As a result, the point instantly returns to normal from a triggered state.

The following record types can be associated to a watchdog record:

• Analog point

• Status point

• Multistate point

• Rate point

• RTString point

• Tank point

• FlowTotals point

If any of the monitored points that are configured for a watchdog record are stale, then the watchdog record will be in an error state. If this error state occurs, then the watchdog record will act as though it were disabled and the associated points will ignore it.