Grafana and Elasticsearch
Optional integrations for run history and metrics.
Requirements for Grafana and Elasticsearch
These integrations are optional. The dashboard runs and executes chaos scenarios without them.
Elasticsearch
The dashboard can connect to an Elasticsearch (or OpenSearch) instance to fetch and display past run details.
- What you need:
- An Elasticsearch or OpenSearch instance reachable from the machine where the dashboard server runs.
- Index name(s) that contain your Krkn run data.
- Credentials: Username and password if the cluster uses basic auth.
Once connected to your instance, you can filter your runs by several different parameters. Clicking a run displays analytics including cluster configurations, node summary information, pod recovery analytics, and object counts.
Grafana
Grafana is used to open pre-built dashboards for a specific chaos run. Grafana dashboards are routed through krkn visualize. The Krkn dashboard uses the parameters of each run to generate a link which returns a Grafana dashboard with a breakdown of that scenario’s data.
- What you need:
- A Grafana instance base URL, connected through krkn visualize.