Modern microservice architectures generate millions of network flows daily. Without deep visibility, teams struggle with security blind spots, performance bottlenecks, and compliance gaps.
No visibility into east-west traffic. Policy misconfigurations go undetected until a breach occurs.
Mean time to resolution measured in hours. Engineers manually correlate logs, metrics, and traces across services.
No automated audit trail for network policies. Manual compliance checks are error-prone and incomplete.
Cilium Vision transforms Cilium's eBPF data plane into an intelligent observability and operations console with 60+ features across security, observability, and operations.
Built on Rust for performance, React for UX, and Cilium eBPF for kernel-level observability.
Every packet traversing the cluster is observed, classified, and enriched with Kubernetes metadata in real-time via Hubble and eBPF.
Every flow enriched with source/destination pod, namespace, labels, verdict, and protocol details.
Select any flow and get a plain-English explanation of what happened, why, and how to fix it.
Record traffic sessions and replay with VCR controls. Jump to any point in time to debug issues.
Automated policy management from discovery to enforcement, with ML-powered confidence scoring and visual rule building.
ML-powered traffic analysis learns communication patterns and generates least-privilege CiliumNetworkPolicies with confidence scores. Observe traffic, generate policies, simulate impact, apply with one click.
Form-based policy creation without YAML. Select namespaces, pods via labels, set ingress/egress rules, ports, CIDRs, FQDNs. Live YAML preview auto-generates as you configure.
Dry-run policy changes before applying. See which flows would be affected, how many services impacted, and get a risk assessment score. No production impact.
Automated compliance checks against CIS, NIST, and SOC2 frameworks. Security posture scoring, findings dashboard, and audit log for every policy change.
From DNS queries to service dependencies, see everything happening in your cluster network.
Interactive topology showing service-to-service communication with traffic volume and latency overlays.
Namespace-to-namespace traffic heatmap revealing communication patterns and unexpected traffic flows.
Real-time DNS query tracking with resolution times, failure rates, and NXDOMAIN detection.
Per-service latency breakdown with P50/P95/P99 percentiles and trend analysis over time.
Network cost breakdown by namespace showing bandwidth consumption and cross-zone traffic charges.
Predictive analytics for traffic patterns, capacity planning, and anomaly prediction.
Automated problem detection, root cause analysis, and remediation. The network fixes itself.
Continuously scans for DNS failures, MTU mismatches, policy denials, and unhealthy endpoints. Proposes and applies fixes automatically with severity-based prioritization.
Correlates packet drops with policies, identifies the exact rule causing denials, and generates one-click fix actions. Reduces MTTR from hours to seconds.
Inject packet loss, latency, and DNS failures via tc-netem. Built-in circuit breaker, preset experiments, and detailed impact metrics for resilience testing.
Progressive traffic shifting with health gates. Promote or rollback canaries based on error rate, latency, and custom SLO thresholds.
Complete visibility and control over every networking layer in your Kubernetes infrastructure.
Service backend health, connection distribution, and failover status.
Route configuration, TLS termination, and traffic routing rules.
IP pool utilization, allocation tracking, and subnet management.
WireGuard/IPsec status, peer connectivity, and cipher configuration.
Multi-cluster connectivity, peer status, and cross-cluster policies.
BGP session status, route advertisements, and peering health.
Outbound policies, SNAT rules, and external connectivity.
Network interface stats, MTU, speed, and error counters.
Everything operators need for production management, from diagnostics to incident response.
Every component chosen for production reliability, speed, and developer experience.
Single binary or containerized. Works with any K3s, EKS, GKE, or AKS cluster running Cilium.