Kubernetes (Helm)

Install the UptimeDuck agent in your cluster with Helm.

Kubernetes monitoring is in Beta and is available on paid plans. Cluster and node limits per plan are on the pricing page.

Helm repository


Repository URL	`https://uptimeduck.com/charts`
Index	`https://uptimeduck.com/charts/index.yaml`
Chart	`uptimeduck-agent`
Latest chart version	`1.0.1` (agent `0.2.0`)
Requirements	Helm 3+

1. Get a cluster token

In panel.uptimeduck.com, open the Kubernetes section and choose Connect Kubernetes Cluster.
Enter a cluster name and, optionally, a project.
The next step, Install with Helm, shows the commands with your token filled in.

Helm is the recommended installation method. Until the cluster reports, its dashboard shows the same Helm commands under Waiting for Cluster Data...

2. Install

helm repo add uptimeduck https://uptimeduck.com/charts
helm repo update
helm install uptimeduck-agent uptimeduck/uptimeduck-agent \
  --set config.token=YOUR_CLUSTER_TOKEN

Override any chart value with additional --set key=value arguments, for example:

helm install uptimeduck-agent uptimeduck/uptimeduck-agent \
  --set config.token=YOUR_CLUSTER_TOKEN \
  --set config.interval=120

Chart values

Value	Default	Description
`config.token`	`""`	Cluster token from the panel. Required. Stored in a Kubernetes Secret.
`config.serverUrl`	`https://api.uptimeduck.com`	Base URL of the UptimeDuck API.
`config.interval`	`60`	Check-in interval in seconds (minimum 60).
`config.autoUpdate`	`false`	Self-update is off in Kubernetes; update with Helm.
`config.allowRemoteConfig`	`true`	Let settings be changed from the panel.
`config.dockerEnabled`	`false`	Docker monitoring plugin.
`config.kubernetesEnabled`	`true`	Kubernetes monitoring plugin.
`image.repository`	`uptimeduck/monitoring-agent`	Agent container image.
`image.tag`	`kubernetes`	Image tag.
`image.pullPolicy`	`Always`	Image pull policy.
`replicaCount`	`1`	Number of agent pods.
`serviceAccount.create`	`true`	Create a ServiceAccount for the agent.
`serviceAccount.name`	`""`	ServiceAccount name; generated when empty.
`rbac.create`	`true`	Create the read-only ClusterRole and ClusterRoleBinding.
`resources`	requests 100m CPU / 128Mi; limits 200m CPU / 256Mi	Pod resources.
`nodeSelector`, `tolerations`, `affinity`	empty	Pod scheduling.
`podAnnotations`, `podSecurityContext`, `securityContext`	empty	Pod metadata and security settings.
`imagePullSecrets`, `nameOverride`, `fullnameOverride`	empty	Standard chart options.

What the chart creates

A Deployment running the agent with -kubernetes-enabled set from config.kubernetesEnabled.
A Secret (name ending in -token) holding config.token, passed to the agent as UPTIMEDUCK_TOKEN.
A ServiceAccount, plus a ClusterRole and ClusterRoleBinding with read-only access:

API group	Resources	Verbs
core	nodes, pods, services, namespaces, events, persistentvolumeclaims, persistentvolumes	get, list, watch
`apps`	deployments, statefulsets, daemonsets, replicasets	get, list, watch
`networking.k8s.io`	ingresses	get, list, watch
`metrics.k8s.io`	nodes, pods	get, list

3. Verify

After installing, the panel shows Waiting for Cluster Data... until the agent reports, then Cluster Connected! and opens the cluster dashboard.