How alarms work
An alarm tells the operator that a signal has reached an abnormal condition and needs attention. The YTCC alarm service watches the signals that have alarm definitions, raises an alarm when a condition is met and clears it when the condition goes away.
Priorities
Every alarm has one of four priorities. Each priority has its own color, shape and number, so it can be told apart without relying on color:
| Number | Priority | Symbol |
|---|---|---|
| 1 | Critical | Red circle |
| 2 | High | Orange triangle |
| 3 | Medium | Yellow square |
| 4 | Low | Purple diamond |
States
| State | Meaning |
|---|---|
| Active (unacked) | The condition is present and nobody has acknowledged the alarm. |
| Acknowledged | The condition is present and the alarm has been acknowledged. |
| Gone (unacked) | The condition has cleared but nobody has acknowledged the alarm. It stays in the list until someone does. |
| Closed | The condition has cleared and the alarm is acknowledged. It leaves the list of active alarms. |
Shelved alarms, alarms out of service and suppressed alarms are not shown in the default view and are not counted or sounded.
Blinking
The priority symbol and the left edge of the row blink; the text never does. Unacknowledged rows are shown in bold.
- Fast blinking: the alarm has come and is not acknowledged.
- Slow blinking: the alarm has gone and is not acknowledged.
- Steady: the alarm is acknowledged.
Alarm levels
An alarm definition can watch a signal with several levels. Four levels are always available: high-high (HH), high (H), low (L) and low-low (LL). You can add above and below levels, a rate of change level, which watches how fast the value changes per second, and a deviation from level, which watches the distance from a setpoint signal. A signal can have at most 32 levels in total.
Acknowledge modes
Each alarm definition sets how its alarms are acknowledged:
| Mode | Behavior |
|---|---|
came (default) | The operator acknowledges the alarm once. An alarm that clears before it is acknowledged stays in the list until it is. |
pair | The arrival and the clearing are acknowledged separately. An acknowledged alarm that clears waits for a second acknowledgement. |
none | The alarm cannot be acknowledged. It starts as acknowledged and disappears when the condition clears. |
deletes | Acknowledging closes the alarm at once, even if the condition is still present. |
Quality and communication alarms
The alarm service evaluates a value only when its quality can be trusted (see Quality). While a signal has no trusted value, its alarms keep their state.
A lost connection raises its own alarm: one per device and one per driver, never one per signal. These communication
alarms have High priority, use the came acknowledge mode and cannot be shelved. They are raised after the
connection has been lost for 5 seconds, and after the alarm service starts it waits 30 seconds before raising a communication alarm.
When the server restarts
Alarms are recorded in the project's history database. When the alarm service starts, it restores the alarms of the last 30 days that were still open, with their state, their times and who acknowledged them; the text and priority of a signal alarm come from its current definition. Live values then update their state. Shelving is restored from the long-term alarm archive. See Alarm archive and restore after a restart.