Kiali — Service Mesh Observability Console for Istio

Introduction

Kiali is the management console for the Istio service mesh. It provides a real-time visual graph of your microservices topology, showing how services communicate, where traffic flows, and where errors occur. Beyond visualization, Kiali validates Istio configuration, detects misconfigurations before they cause outages, and integrates with Prometheus, Grafana, and Jaeger to provide a unified observability experience for mesh-managed applications.

What Kiali Does

• Renders an interactive service graph showing real-time traffic flows between microservices
• Validates Istio configuration objects and flags misconfigurations or conflicts
• Displays per-service metrics including request rates, error rates, and latency percentiles
• Integrates distributed tracing from Jaeger or Zipkin to show request paths across services
• Provides wizards for creating traffic routing rules, circuit breakers, and fault injection

Architecture Overview

Kiali runs as a single Go binary deployed in the istio-system namespace. It queries the Kubernetes API for workloads, services, and Istio custom resources (VirtualServices, DestinationRules, etc.). It pulls metrics from Prometheus to compute traffic rates and health indicators. The frontend is a React single-page application that renders the service topology graph using the Cytoscape.js library. Kiali does not sit in the data path and adds no latency to service-to-service communication.

Self-Hosting & Configuration

• Deploy as an Istio addon with a single kubectl apply or install via the Kiali Helm chart
• Configure the Kiali CR (Custom Resource) to set authentication mode (token, OpenID Connect, anonymous)
• Point Kiali to your Prometheus and Grafana instances via the external_services configuration
• Enable Jaeger or Tempo integration for distributed tracing links in the service graph
• Use namespace filtering to scope the dashboard to specific teams or application namespaces

Key Features

• Real-time service topology graph with animated traffic flow and health color coding
• Istio configuration validation that catches 50+ types of misconfigurations automatically
• Traffic routing wizards for creating A/B tests, canary deployments, and traffic mirroring
• Health dashboards showing request success rates, throughput, and latency per service
• Multi-cluster support for visualizing service mesh topology across federated Istio installations

Comparison with Similar Tools

• Grafana + Prometheus — excellent for metrics dashboards but lacks service topology visualization and Istio config validation
• Jaeger — focused on distributed tracing, does not show mesh topology or validate Istio resources
• Linkerd Viz — built-in dashboard for Linkerd mesh, not compatible with Istio
• Meshery — multi-mesh management tool but broader scope with less depth in Istio-specific observability
• Backstage — developer portal with service catalog but no real-time traffic visualization or mesh integration

FAQ

Q: Does Kiali work without Istio? A: Kiali is designed specifically for Istio. It requires Istio sidecar proxies to generate the telemetry data it visualizes. It does not work with other service meshes.

Q: Does Kiali add latency to my services? A: No. Kiali is a control-plane tool that reads metrics from Prometheus and configuration from the Kubernetes API. It never intercepts or proxies service traffic.

Q: Can I use Kiali with Tempo instead of Jaeger? A: Yes. Kiali supports Grafana Tempo as a tracing backend. Configure the tracing URL in the Kiali CR to point to your Tempo query frontend.

Q: Is Kiali production-ready? A: Yes. Kiali is the official Istio observability console, maintained by Red Hat, and is widely used in production Istio deployments across enterprises.

Sources

• https://github.com/kiali/kiali
• https://kiali.io/docs