Summary

kwatch monitors Kubernetes clusters, instantly detecting and reporting failures with actionable insights delivered to team chat. It targets users seeking immediate, human-readable alerts and diagnostic hints, eliminating the need for complex backends or dashboards. The core benefit is reduced Mean Time To Resolution (MTTR) through clear, root-cause-focused failure explanations.

How It Works

The system builds a dependency graph of Kubernetes resources from informer caches, mapping pods to nodes, owners, services, and more. When an incident occurs, kwatch analyzes it against this graph to identify root causes and impact. It features mass failure detection for widespread issues stemming from shared dependencies and offers an optional, opt-in AI component that analyzes logs within the cluster to suggest remediation steps. This approach prioritizes actionable alerts over raw data aggregation.

Quick Start & Requirements

Installation is streamlined via Helm (helm install kwatch kwatch/kwatch --version 0.11.0) or kubectl apply using a fetched configuration. kwatch is a single binary, requiring only a Kubernetes cluster and no external dependencies or persistent storage. Setup is estimated at approximately 5 minutes.

Highlighted Details

• Monitors over 20 Kubernetes signals by default, including pod crashes (OOMKilled, CrashLoopBackOff), pending pods, node issues, PVC space, failed jobs, stuck rollouts, and control-plane health.
• Context-aware intelligence correlates incidents with resource dependencies (nodes, ConfigMaps, Secrets) to pinpoint root causes and impact.
• Mass failure detection identifies widespread issues by analyzing shared dependencies among failing resources.
• Optional AI-powered troubleshooting analyzes logs for likely causes and remediation steps, running entirely within the cluster.

Maintenance & Community

The project encourages community engagement via a Discord channel for questions and suggestions. It lists "Contributors" and "Stargazers," indicating active community interest.

Licensing & Compatibility

kwatch is released under the permissive MIT License, allowing unrestricted use, modification, and distribution, including in commercial and closed-source environments.

Limitations & Caveats

kwatch is explicitly not a full observability platform; it does not provide metrics collection, dashboards, or log storage. The AI-powered troubleshooting feature is limited to linux/amd64 and linux/arm64 platforms. TLS certificate monitoring is disabled by default.