Indicator Type | Budgeting Method | Content of Queried Metric Data | Time Grouping | Examples of the Data |
---|---|---|---|---|
thresholdMetric | Occurrences | each occurence of logged indicator values | none | Request latency |
thresholdMetric | TimeSlices | each occurence of logged indicator values | on timeSliceWindow boundaries | Request latency |
ratioMetric | Occurrences | each occurence of logged indicator values¹ | none¹ | http 200 requests vs all http requests |
ratioMetric | TimeSlices | each occurence of logged indicator values | on timeSliceWindow boundaries | http 200 requests vs all http requests |
??? | ??? | each occurence of evaluated slo met/missed bool values² | none² | slo met/missed (bool) for request latency on 5m windows; Scenario 1, TimeSlice based EB – Result Row |
??? | ??? | each occurence of evaluated slo window performance float values² | none² | slo window performance (float) for request latency on 5m windows; Scenario 1, Weighted TimeSlice based EB – Result Row |
¹ Data may already be time–grouped within the metric itself.
² Data is already time–grouped within the metric itself.
- Content of Queried Metric Data – what data the metric referenced in query actually contains.
- Time Grouping – any time grouping that should be done when evaluating the SLO that's not already baked into the metric itself or the query