connected-home/helm/redpanda-operator

Redpanda Operator

Installation

Prerequisite

To deploy operator with webhooks (enabled by default) please install cert manager. Please follow the installation guide

The cert manager needs around 1 minute to be ready. The helm chart will create Issuer and Certificate custom resource. The webhook of cert-manager will prevent from creating mentioned resources. To verify that cert manager is ready please follow the verifying the installation

The operator by default exposes metrics endpoint. By leveraging prometheus operator ServiceMonitor custom resource metrics can be automatically discovered.

1. Install Redpanda operator CRDs:

kubectl apply -k 'https://github.com/redpanda-data/redpanda/src/go/k8s/config/crd?ref=v21.3.4'

The CRDs are decoupled from helm chart, so that helm release can be removed without cascading deletion of underling custom resources. Other argument for decoupling is that helm cli can incorrectly patch the Custom Resource Definition.

Helm installation

1. Install the Redpanda operator:

The example command should be invoked from src/go/k8s/helm-chart/charts

helm install --namespace redpanda-system --create-namespace redpanda-system ./redpanda-operator

Alternative installation with kube-prometheus-stack that includes prometheus operator CRD

helm install --dependency-update \
--namespace redpanda-system \
--set monitoring.enabled=true \
--create-namespace redpanda-operator ./redpanda-operator

Other instruction will be visible after installation.

Values

Key	Type	Default	Description
affinity	object	{}	Allows to specify affinity for Redpanda Operator PODs
config.apiVersion	string	"controller-runtime.sigs.k8s.io/v1alpha1"
config.health.healthProbeBindAddress	string	":8081"
config.kind	string	"ControllerManagerConfig"
config.leaderElection.leaderElect	bool	true
config.leaderElection.resourceName	string	"aa9fc693.vectorized.io"
config.metrics.bindAddress	string	"127.0.0.1:8080"
config.webhook.port	int	9443
configurator.pullPolicy	string	"IfNotPresent"	Define the pullPolicy for Redpanda configurator image
configurator.repository	string	"vectorized/configurator"	Repository that Redpanda configurator image is available
configurator.tag	string	"{{ .Chart.AppVersion }}"	Define the Redpanda configurator container tag
clusterDomain	string	cluster.local	Defines Kubernetes Cluster Domain
fullnameOverride	string	""	Override the fully qualified app name
image.pullPolicy	string	"IfNotPresent"	Define the pullPolicy for Redpanda Operator image
image.repository	string	"vectorized/redpanda-operator"	Repository that Redpanda Operator image is available
image.tag	string	"{{ .Chart.AppVersion }}"	Define the Redpanda Operator container tag
imagePullSecrets	list	[]	Redpanda Operator container registry pullSecret (ex: specify docker registry credentials)
labels	string	nil	Allows to assign labels to the resources created by this helm chart
logLevel	string	"info"	Set Redpanda Operator log level (debug, info, error, panic, fatal)
monitoring	object	{"enabled":false}	Add service monitor to the deployment
nameOverride	string	""	Override name of app
nodeSelector	object	{}	Allows to schedule Redpanda Operator on specific nodes
podAnnotations	object	{}	Allows setting additional annotations for Redpanda Operator PODs
podLabels	object	{}	Allows setting additional labels for for Redpanda Operator PODs
rbac.create	bool	true	Specifies whether the RBAC resources should be created
replicaCount	int	1	Number of instances of Redpanda Operator
resources	object	{}	Set resources requests/limits for Redpanda Operator PODs
serviceAccount.create	bool	true	Specifies whether a service account should be created
serviceAccount.name	string	nil	The name of the service account to use. If not set name is generated using the fullname template
tolerations	list	[]	Allows to schedule Redpanda Operator on tainted nodes
webhook.enabled	bool	true