CoreOS etcd-operator

etcd-operator Simplify etcd cluster configuration and management.

DISCLAIMER: While this chart has been well-tested, the etcd-operator is still currently in beta. Current project status is available here.

Introduction

This chart bootstraps an etcd-operator and allows the deployment of etcd-cluster(s).

Official Documentation

Official project documentation found here

Prerequisites

• Kubernetes 1.4+ with Beta APIs enabled
• Suggested: PV provisioner support in the underlying infrastructure to support backups

Installing the Chart

To install the chart with the release name my-release:

$ helm install stable/etcd-operator --name my-release

Note: If you set cluster.enabled on install, it will have no effect. Before you create an etcd cluster, the TPR must be installed by the operator, so this option is ignored during helm installs, but can be used in upgrades.

Uninstalling the Chart

To uninstall/delete the my-release deployment:

$ helm delete my-release

The command removes all the Kubernetes components EXCEPT the persistent volume.

Updating

Updating the TPR resource will not result in the cluster being update until kubectl apply for TPRs is fixed see kubernetes/issues/29542 Work around options are documented here

Configuration

The following tables lists the configurable parameters of the etcd-operator chart and their default values.

Parameter	Description	Default
replicaCount	Number of etcd-operator replicas to create (only 1 is supported)	1
image.repository	etcd-operator container image	quay.io/coreos/etcd-operator
image.tag	etcd-operator container image tag	v0.3.2
image.pullPolicy	etcd-operator container image pull policy	IfNotPresent
resources.limits.cpu	CPU limit per etcd-operator pod	100m
resources.limits.memory	Memory limit per etcd-operator pod	128Mi
resources.requests.cpu	CPU request per etcd-operator pod	100m
resources.requests.memory	Memory request per etcd-operator pod	128Mi
nodeSelector	node labels for etcd-operator pod assignment	{}
cluster.enabled	Whether to enable provisioning of an etcd-cluster	false
cluster.name	etcd cluster name	etcd-cluster
cluster.version	etcd cluster version	v3.1.8
cluster.size	etcd cluster size	3
cluster.backup.enabled	Whether to create PV for cluster backups	false
cluster.backup.provisioner	Which PV provisioner to use	kubernetes.io/gce-pd (kubernetes.io/aws-ebs)
cluster.backup.config.snapshotIntervalInSecond	etcd snapshot interval in seconds	30
cluster.backup.config.maxSnapshot	maximum number of snapshots to keep	5
cluster.backup.config.storageType	Type of storage to provision	PersistentVolume
cluster.backup.config.pv.volumeSizeInMB	size of backup PV	512MB
cluster.restore.enabled	Whether to restore from PV	false
cluster.restore.config.storageType	Type of storage to restore from	PersistentVolume
cluster.restore.config.backupClusterName	Name of cluster to restore from	etcd-cluster
cluster.pod.antiAffinity	Whether etcd cluster pods should have an antiAffinity	false
cluster.pod.resources.limits.cpu	CPU limit per etcd cluster pod	100m
cluster.pod.resources.limits.memory	Memory limit per etcd cluster pod	128Mi
cluster.pod.resources.requests.cpu	CPU request per etcd cluster pod	100m
cluster.pod.resources.requests.memory	Memory request per etcd cluster pod	128Mi
cluster.pod.nodeSelector	node labels for etcd cluster pod assignment	{}
rbac.install	install required rbac service account, roles and rolebindings	false
rbac.apiVersion	rbac api version `v1alpha1	v1beta1`

Specify each parameter using the --set key=value[,key=value] argument to helm install. For example:

$ helm install --name my-release --set image.tag=v0.2.1 stable/etcd-operator

Alternatively, a YAML file that specifies the values for the parameters can be provided while installing the chart. For example:

$ helm install --name my-release --values values.yaml stable/etcd-operator

RBAC

By default the chart will not install the recommended RBAC roles and rolebindings.

To determine if your cluster supports this running the following:

$ kubectl api-versions | grep rbac

You also need to have the following parameter on the api server. See the following document for how to enable RBAC

--authorization-mode=RBAC

If the output contains "beta" or both "alpha" and "beta" you can may install with enabling the creating of rbac resources (see below).

Enable RBAC role/rolebinding creation

To enable the creation of RBAC resources (On clusters with RBAC). Do the following:

$ helm install --name my-release stable/etcd-operator --set rbac.install=true

Changing RBAC manifest apiVersion

By default the RBAC resources are generated with the "v1beta1" apiVersion. To use "v1alpha1" do the following:

$ helm install --name my-release stable/etcd-operator --set rbac.install=true,rbac.apiVersion=v1alpha1