With Monitis transition to the new Alerts 2.0 alerting model the Chart and Table Views of the migrated monitors (see Migration to Alerts 2.0) undergo the following changes.
The Chart and Table View will now reflect changes in the monitor’s state.
1. CHECK FAILED
If you do not have any threshold set on your monitor and the check fails (result is error), the monitor will enter the CHECK FAILED state.
In the chart view the checks that have returned CHECK FAILED state will be denoted by solid grey dots.
In the table view the checks that have returned Check Failed state will be denoted by a grey “X” icon.
2. User-defined states: CRITICAL and WARNING
If you have thresholds set up in your monitor (see Thresholds), it will enter the CRITICAL or WARNING state set in the threshold as soon as the threshold condition is matched.
In the Chart View the checks that have returned WARNING or CRITICAL state will be highlighted with yellow and red hatched bars respectively.
In the Table View the checks that have returned WARNING or CRITICAL state will be marked with yellow “i” and red “X” icons respectively.
3. THRESHOLD NOTE
There can be cases that the monitor hasn’t entered WARNING or CRITICAL state yet (threshold not violated as the set count of consecutive checks not reached yet) but the availability or performance part of the monitor’s threshold (see Thresholds) has been violated.
We call this state “THRESHOLD NOTE”. For example, if the threshold’s condition is set to “Monitor will enter the Critical state if for 5 consecutive checks: Result is error or ANY metric matches the condition below”, when this condition is matched for the first 4 consecutive checks the threshold is not yet violated as the count of 5 is not reached yet.
For the first 4 failed checks the THRESHOLD NOTE state will be shown in monitor charts and tables as follows:
In the Chart View the check will not have any special marking on it, but if you roll a mouse cursor over it you will see the THRESHOLD NOTE state data in the tooltip.
In the Table View the check will be marked with a small blue exclamation sign “!” on the OK icon.
As soon as the monitor fails for the 5th consecutive time, the status will change from THRESHOLD NOTE to CRITICAL.
4. NO THRESHOLD
If there is no threshold set in your monitor, the monitor state is NO THRESHOLD, shown with grey “i” icon in the table view.
5. INSUFFICIENT DATA
There may be cases that some of the monitored metric(s) do not report any value during the check, e.g. if there was some problem with the monitored object.
If the empty metric is part of the metric expression in the threshold (see Thresholds), then there may be cases that the expression cannot be resolved.
For example, in the shown threshold the metric expression consists of 2 metrics tied with the logical AND operator.
If during some check the value reported by one of the metrics crosses the threshold, while the other metric doesn’t report any value, the AND expression cannot be resolved, as its result depends on the missing value of the second metric.
As you can see it on the below screen, during the last 3 checks Metric A has reported values which are all above the set threshold of “Metric A > 1”: 48, 55, and 88.
At the same time, during all these 3 checks metric B hasn’t reported any value.
The AND expression for all these checks cannot be resolved without the value of Metric B, which is missing. That’s why the monitor will enter a state called INSUFFICIENT DATA, marked in the table view with a grey icon with question sign “?”.
Note that in the chart view whenever a metric doesn’t report any value there will be a gap in the metric’s line.
For Aggregated Data
When your monitor shows aggregated data (see Data Aggregation in Monitors), you will see also the % breakdown of monitor states per the aggregation intervals in chart and table views. E.g. on the below screen you can see in the table view % breakdown of monitor states for each one-hour aggregation interval.
And e.g. here you can see in the chart view % breakdown of monitor states for each one-day aggregation interval.
The % of monitor states within the reporting interval is calculated as follows:
TOTAL = OK+NOK+WARNING+THRESHOLD NOTE+INSUFFICIENT DATA
UPTIME (%) = ((OK+THRESHOLD NOTE)/TOTAL)*100%
NOK (%) = (NOK/TOTAL)*100%
WARNING = (WARNING/TOTAL)*100%
INSUFFICIENT DATA = (INSUFFICIENT DATA/TOTAL)*100%