Kubernetes monitoring made simple
Stop drowning in pod-level noise. Sentinel monitors the health of your Kubernetes workloads — Deployments, StatefulSets and DaemonSets — and alerts you the moment something breaks. Install the operator once, annotate your workloads, done.
How Kubernetes monitoring works
Three steps. No sidecars, no scraping, no Prometheus to babysit.
Deploy the operator
Install the Sentinel operator into your cluster with a single Helm command. It watches the Kubernetes API directly.
Annotate your workloads
Add one annotation to the Deployments or StatefulSets you want to track. The operator picks them up automatically.
Get alerted
Sentinel reports Up, Degraded or Down based on ready vs desired replicas and notifies you on Slack, email or webhook.
Native Kubernetes Monitoring
Deploy the operator once. Annotate your workloads. Sentinel takes care of the rest.
Workload health, not pod noise
Monitors Deployments, StatefulSets and DaemonSets directly. Reports health based on ready vs desired replicas — not individual pod events.
Rolling update detection
Knows when a rollout is in progress and reports it as metadata. No false alerts during deploys.
Three health states
Up (all pods ready), Degraded (partial availability), Down (no pods ready). Tracked over time with uptime history.
Multi-component services
Multiple workloads with the same service slug are tracked as components. Expand the uptime bar to see per-component health breakdown.
Install the operator
helm install sentinel-operator \
oci://ghcr.io/sentinel-app-io/charts/sentinel-operator \
--namespace sentinel --create-namespace \
--set credentials.sentinelUrl=https://app.sentinel-app.io \
--set credentials.keyId=your-key-id \
--set credentials.secret=your-secretAnnotate your workloads
apiVersion: apps/v1
kind: Deployment
metadata:
name: payment-api
annotations:
sentinel.io/monitor: "true"
sentinel.io/service: "payment-api"
# sentinel.io/component: "api" # optional overrideKubernetes monitoring FAQ
How do I monitor a Kubernetes deployment?
Install the Sentinel operator in your cluster with Helm, then add the annotation sentinel.io/monitor: "true" to any Deployment, StatefulSet or DaemonSet. Sentinel automatically tracks its health based on ready versus desired replicas and alerts you when it degrades.
What is the difference between pod monitoring and workload monitoring?
Pod-level monitoring floods you with events every time a pod restarts or reschedules. Sentinel monitors workload health instead — it reports a service as down only when the workload as a whole can't serve, not on individual pod churn. Far fewer false alerts.
Does Sentinel work with EKS, GKE and AKS?
Yes. The operator talks to the standard Kubernetes API, so it runs on any conformant cluster — Amazon EKS, Google GKE, Azure AKS, k3s or self-hosted. No cloud-specific configuration required.
Does monitoring during a rolling update trigger false alerts?
No. Sentinel detects when a rollout is in progress and reports it as metadata rather than an outage, so deploys don't page your on-call.