Combine Threshold Monitoring with Data Monitoring
When creating a new Alarm you must enable one or more of the 3 independent alarm types:
- Threshold
- Health
- Data Monitoring
There does not seem to be any logical reason that when an alarm should monitor Thresholds that it shouldn't also be monitoring Data and vice versa.
We ran into an issue when we built some DB Query Monitors and attached them to Alarms that didn't have Threshold monitoring enabled, only Data Monitoring. We never received alerts when the DB Query was triggering a failure. Only upon detailed investigation when we noticed that we were in a failure condition but had never received an email did we track down that Threshold monitoring was not enabled and DB Queries are part of Threshold, not Data Monitoring.
-
Craig Kosior commented
Basically, I would want to revamp your Alarm creation/edit screens to make alarm creation / configuration much simpler
At a high level:
Current system:
- Threshold
- Health Check
- Data Monitor
- User needs to make sure they have the proper alarm type enabled or they may not get emails when an alarm is triggered.
- It is not clear that selecting Health Check means getting the daily/weekly email showing the current status.Future system:
- Monitor
- Status Check
- Turning on monitor would mean that when any style of monitoring is triggered/violated, an email is sent and continued to be sent if it is a threshold violation until cleared or max number reached
- Turning on Status Check would mean that an email would be sent showing the overall status on the schedule chosen -
Craig Kosior commented
I do understand that Health monitoring is independent of the other two in that any alarm that is a healh alarm sends an email showing the status of everything in that alarm on the schedule provided. Therefore, Health monitoring should be renamed to something else which clearly explains that enabling this will send status checks on the scheduled times of the alarm.
Perhaps having a checkbox saying 'Send Scheduled Health Checks on this Alarm instead of 'Health Check Alarm' would be clearer.