we are looking for a replacement of currently used load running framework and we are evaluating K6 and Locust. If you are interested in more details see the following discussion.
I would like to ask a few questions about K6 support for distributed runs.
There is a support for HTML reports for single run, is it possible to have the same for distributed runs?
For single run it is possible to output to json using --out json=file.json, is there a way to merge those files from distributed run and generate report from it?
K6 operator supports running distributed runs (it makes sure the runs are synchronized) is there a way to the the same outside of Kubernetes, for example, cloud providers AWS/Azure or even bare metal?
Hello! I am glad you are considering using k6 to replace your existing framework.
Regarding (1)/(2): We recommend sending the metrics to a central place (e.g., a Prometheus instance) and then using Grafana to visualize/aggregate them. There is a dashboard export feature in Grafana that can be used to generate PDF reports.
Regarding (3): Sadly, we don’t provide non-K8s support. You could run the operator in the managed k8s offerings of AWS/Azure or go DIY, but I understand both are less than ideal.
That said, we have a program that offers a Grafana Cloud Pro subscription (free of charge) to open-source projects—Keycloack, being part of the CNCF foundation, would qualify!
With the Cloud offering of k6, you get (1)/(2) out of the box, and regarding (3), we have a few options that maybe could solve your pain (e.g., we can run the tests in our managed load gens). If you’d like to discuss this further or set up a call, feel free to email me or reach out in our Community Slack.
Cheers!
Daniel // dgzlopes at grafana dot com // dgzlopes (Slack/GitHub)
