Properly document output (i.e. collector) options #743
Comments
Now buckets with fewer HTTP samples can also be aggregated, with outlier detection using the original sorting method, but with interpolation for improved accruacy. This is a more resource-intensive process, so for larger sample sizes quickselect is still the preferred algorithm. There's also a new AggregationSkipOutlierDetection option that disables outlier detection altogether. It is disabled by default and is intended mostly for testing purposes. And finally, the in-code documentation about the aggregation algorithm and options is expanded, which should help with https://github.com/loadimpact/k6/issues/628 and #743
|
We probably should split up the results output docs page into multiple pages, one summary page that has an overview and explains how metric outputs work, and for each different output. The current single page is getting huge and it's proving insufficient to properly document the end-of-test summary and all of the supported external outputs and their options. As is evident from this issue and others, we need to better document the InfluxDB output and its options. For it, we might also document common issues that people run into, since there are a few issues people run into fairly frequently. https://github.com/loadimpact/k6/issues/628 is another connected issue to improve the docs about the cloud collector options. And depending on when and how we implement #587, we may need to have a fairly comprehensive section with deprecation information about the environment variable options.... |
|
Closing this, since this was mostly implemented in the new docs at https://k6.io/docs/, and the remaining things are separate issues in the new docs repo (e.g. grafana/k6-docs#179, grafana/k6-docs#180) |
Specifically, the different environment variables available. Also, after #587 is done, the JS/JSON options as well.
The text was updated successfully, but these errors were encountered: