KlusterAlert
Overview
KlusterAlert monitors your Kubernetes workloads, evaluates alert rules as code, and notifies the right people through Slack, Microsoft Teams, Google Chat, PagerDuty, email, or any custom webhook — before users notice a problem.
How it works
1
Deploy the agent
Install the KlusterAlert Helm chart into each cluster. The lightweight agent runs in its own namespace and requires minimal RBAC.
2
Agent streams events
The agent watches pod events, node metrics, and resource usage. It streams events to the KlusterAlert ingestion API.
3
Rules are evaluated
Your alert rules run against the incoming event stream. Conditions are evaluated in sliding time windows.
4
Notifications fire
When a condition triggers, KlusterAlert sends a notification to your configured channels with full context and runbook link.
Key concepts
Clusters
Each Kubernetes cluster registers with KlusterAlert via the Helm chart. You can monitor unlimited clusters from one account.
Alert Rules
YAML files that define conditions, time windows, severity, and notification targets. Version them in Git alongside your K8s manifests.
Notification Channels
Slack, Microsoft Teams, Google Chat, PagerDuty, email, and custom webhooks. Route alerts by severity to different channels.
Runbooks
Attach a runbook URL or inline markdown to each alert rule. Responders see the resolution steps directly in the notification.
Historical Metrics
90 days of cluster metrics stored. Browse event timelines, correlate incidents with deployments, and understand trends.
Security Checks
Built-in rules surface privileged pods, exposed secrets, and RBAC misconfigurations without extra configuration.
Ready to connect your first cluster?
The quickstart walks you through installing the Helm chart and setting up your first alert rule in under 5 minutes.
Go to Quickstart