The four categories of alarm conditions are color-coded as follows:
- Red
- A condition that requires immediate attention. Unless the problem is
resolved immediately, the system cannot continue to run properly.
- Yellow
- A condition that may require prompt resolution. Such a condition often
means there is a buffer pool or disk space problem. The system can continue
to operate, but performance will suffer.
- Green
- An earlier alarm condition has been cleared. For example, the system
may have generated a red alarm because the CPU was overloaded. Alleviating
that condition produces a green alarm. A green alarm condition requires no
attention on your part (other than deleting the message from the System Monitor).
- White
- An informational message has been written to the error log, recording
a routine system event or a low severity alarm condition. You will probably
generate white alarms during application development, but should aim to ensure
that they won't occur during production use, by testing for the conditions
that cause them in your state tables and custom servers.
When the system monitor window is minimized, the color of the icon indicates
the severest category of alarm outstanding (green being counted as “more
severe” than white).
To assist users with limited color vision, in addition to the color indicator,
text is displayed inside the icon. The text displayed depends, like the color,
on the severity of the alarm:
- A Red alarm is indicated by the text ERROR
- A Yellow alarm is indicated by the text WARN
- A Green alarm is indicated by the text INFO
- A White alarm is indicated by the text NOTIFY
If there is no alarm, there is no text, and the icon background is blue.