The issue is not that Dynatrace cannot chart both values at the same time. The issue is that the values are identical. Let me explain.
Unless configured otherwise in the Code mode tab, the Data Explorer performs a space aggregation on the metric datapoints. After applying such a space aggregation in Dynatrace, you can use all available aggregations on the transformed data. In many cases, this behavior is reasonable, but there are pitfalls.
For example, consider a counter metric that natively only supports the aggregation value. The metrics has the dimension host and the following values per host:
| | timeslot1 |
| HOST-1 | 1 |
| HOST-2 | 5 |
| HOST-3 | 2 |
After applying a space aggregation, the per host values are merged and you can chart the min (= 1), the max (= 5), the sum (= 8), the avg (= 2.66) and so on. Also the percentiles are calculated based on the three values. Reasonable, isn't it?
In contrast, let's assume there is only the series for HOST-1 available and no additional series/dimensions. After applying the space aggregation, you can again apply all aggregations. However, as there is only one series, the space aggregation does in fact not aggregate any values. Consequently, the values for avg, min, max, sum as well as the percentiles are the same (i.e., 1 for the example above).
The issue is also described on the Metrics API FAQ page.
