kspan

kspan

Installs and configures kspan to get tracing from your Kubernetes events

kspan is a library to generate traces from Kubernetes events and sends them to an OpenTelemetry protocol (OTLP) compatible backend.

Kubernetes is an application used to manage your workloads. So let's trace it like an application.

Installing the Chart

You will likely need to configure multiple args and optionally the env parameter as well. These are array style parameters which are best configured within a custom values file. These parameters are used to configure the command-line arguments for kspan, as well as any optional environment variables you may want to set up to pass in sensitive data.

The configuration will be based on the specific OTLP enabled backend that you are using. This example demonstrates how to configure for use with Honeycomb.io by referencing an existing secret for the API key.

args:
  - --otlp-addr=api.honeycomb.io:443
  - --otlp-headers=x-honeycomb-team=$(HONEYCOMB_API_KEY),x-honeycomb-dataset=kubernetes-traces
  - --otlp-secured

env:
  - name: HONEYCOMB_API_KEY
    valueFrom:
      secretKeyRef:
        name: honeycomb
        key: api-key

With a values file specified, you can install the chart

helm repo add puckpuck https://puckpuck.github.io/helm-charts
helm install kspan puckpuck/kspan --values my-values.yaml

Configuration

The values.yaml file contains information about all configuration options for this chart.

Parameters

Parameter	Description	Default
args	Array of command arguments to pass into the kspan process	[]
env	Array of environment variables to set on the kspan pod	[]
image.repository	kspan Image repository	puckpuck/kspan
image.tag	kspan Image tag (leave blank to use app version)	nil
image.pullPolicy	kspan Image pull policy	IfNotPresent
resources	CPU/Memory resource limits/requests	{}
nodeSelector	Node labels for pod assignment	{}
tolerations	Toleration labels for pod assignment	[]
affinity	Affinity settings for pod assignment	{}