Peak is open source software to help you manage the performance of your web applications and APIs.
Peak uses the Kubernetes Batch API to load-test your web apps and APIs. When you create a test suite, you're asked to specify the number of nodes that will be used for the test. Peak will create each node as an Alpine Linux container running in a Kubernetes pod, and each measures the response time of requests and posts the results back to an InfluxDB instance. You can then use Grafana with the Peak dashboard to visualise your results.
Peak provides an Ansible Operator that you can use to easily deploy and consume the application on OpenShift. There's a step-by-step guide to deployment provided in the Peak Operator docs.
If you don't have access to the Operator Framework, Peak also provides an OpenShift template to help you get an up-and-running rapidly.
Firstly, create a new OpenShift project:
oc new-project peak-project
Deploy a PostgreSQL database for the peak API service:
oc new-app postgresql-persistent -p POSTGRESQL_DATABASE=peakdb -p VOLUME_CAPACITY=1Gi -p POSTGRESQL_VERSION=9.6
Create app components from the template:
oc new-app -f <(curl https://raw.githubusercontent.com/peak-oss/peakdocs/master/templates/peak.yaml)
Peak uses a custom Grafana dashboard to display metrics, and the following steps show you how to configure Grafana.
Access the Grafana URL, and enter the default credentials admin/admin when prompted. Create a new, secure password when prompted
Select Add datasource from the home screen, and enter the following:
| Key | Value |
|---|---|
| Name | influxdb |
| Type | InfluxDB |
| URL | http://influxdb:8086 |
| Database | peakdb |
Import a dashboard, and specify 8463 as the Grafana ID. When prompted in the following screen, select the InfluxDB datasource.
That's it! Your peak instance should be able to run tests against your APIs, report the metrics into InfluxDB, and visualise in the Grafana dashboard.