09 Sep 2023
05:18 PM
- last edited on
11 Sep 2023
09:41 AM
by
Ana_Kuzmenchuk
We are integrating data from K6 into Dynatrace using metric API - https://k6.io/docs/results-output/real-time/dynatrace/
K6 is ingesting 6 data points for a minute - but dynatrace is aggregating the data across those 6 data points but they are looking to see how can we not aggregate the data for a minute in Dynatrace
Below table shows they sent 10 data points for a metric
data:{"time":"2023-09-08T10:38:13.2067109-05:00" | value | 67.38 |
data:{"time":"2023-09-08T10:38:23.0113171-05:00" | value | 69.66 |
data:{"time":"2023-09-08T10:38:33.0136266-05:00" | value | 63.24 |
data:{"time":"2023-09-08T10:38:43.0240259-05:00" | value | 63.24 |
data:{"time":"2023-09-08T10:38:53.0269623-05:00" | value | 61.76 |
data:{"time":"2023-09-08T10:39:03.032395-05:00" | value | 59.2205 |
data:{"time":"2023-09-08T10:39:13.0439243-05:00" | value | 60.5196 |
data:{"time":"2023-09-08T10:39:23.0481963-05:00" | value | 63.354 |
data:{"time":"2023-09-08T10:39:33.0678128-05:00" | value | 77.1408 |
data:{"time":"2023-09-08T10:39:43.0538723-05:00" | value | 59.6212 |
But Dynatrace data for every minute, so there is a mismatch in calculation in Dynatrace compared to K6
Test Time | 10:38 - 10:43 | |||
Metric Name | K6 Metrics | Dynatrace Metrics | api/v2/metrics/ingest data | Resolution 5 Mins |
Average | 64.51 | 65.1 | 64.51409 | 64.5 |
Minimum | 59.22 | 61.8 | 59.2205 | 59.2 |
Median | 63.24 | 65.1 | 63.23985 | 64.5 |
Max | 77.14 | 77.1 | 77.1408 | 77.1 |
90th % | 70.41 | 65.1 | 70.40637 | 64.5 |
95% | 73.77 | 65.1 | 73.773585 |
64.5 |
They could use Max, but they are interested in seeing 90th and 95th percentile metrics which is huge difference compared the metrics from K6 calculation
Please provide suggestions
Solved! Go to Solution.
11 Sep 2023 08:39 AM
Dynatrace metrics have the lowest granularity of 1 minute regardless of the ingestion type. You can use the max / min aggregations as you have already mentioned.
What is your source of the metric data? Maybe you can capture the relevant metrics using different means (tracing) and calculate the metrics based on captured trace data.