Skip to main content

Project Overview

What is K8s Sentinel​

K8s Sentinel is a Kubernetes security management platform built on Cilium and Tetragon (eBPF), bringing the capabilities a cluster's security posture needs into one graphical console:

  • Runtime security: monitor and block process executions and file access with TracingPolicy (Monitoring / Protect modes)
  • Network security: manage Pod Ingress / Egress access with Cilium Network Policy, with Network Topology showing actual traffic and the external exposure surface
  • Admission control: intercept non-compliant configuration at creation time with ValidatingAdmissionPolicy
  • Vulnerability scanning: show the known vulnerabilities (CVEs) in the images of running workloads and of registries, scanned by Trivy; combine with Admission Policy to refuse images over a vulnerability limit
  • Incident response: a live security event stream, one-click Pod network isolation (Quarantine), webhook alerts and Syslog forwarding
  • Auditability: every operation and sign-in recorded in the Audit Log

DevSecOps engineers and Platform teams manage the full lifecycle of security policies and see the cluster's security state in real time, without hand-writing YAML or driving everything through kubectl, dramatically lowering the barrier to operating Kubernetes security.

Core Features​

ModuleDescription
TracingPolicy ManagementVisually create, edit, enable, disable, and delete TracingPolicies - no manual YAML required
Admission PolicyGraphically build Kubernetes ValidatingAdmissionPolicies with fourteen rule types: Label, Image, resource limits, hostPath, environment variables, ConfigMap / Secret size, vulnerability limits, and more
Network PolicyGraphically build Cilium Network Policies to manage Pod Ingress / Egress access rules, including L7 HTTP rules
QuarantineOne-click network isolation of a suspicious Pod for incident response - the process and memory stay intact for forensics
Network TopologyInteractive graph of Pod connections observed by Cilium (including L7) - quickly identify anomalous network traffic; Pods with Critical vulnerabilities are marked
Vulnerabilities (Workloads)List the known CVEs in the images of running workloads (read from Trivy Operator's reports), with workloads that are both Critical and externally exposed first; not scanned is never shown as clean
Vulnerabilities (Registries)Scan the images in Harbor, Docker Registry and quay.io before deployment, with Filters choosing the scope, on a schedule or triggered manually
Behavior DiscoveryAutomatically analyze workload behavior in the cluster to help engineers discover security baselines
Security EventsReal-time display of kprobe security events captured by Tetragon, with filtering and tracking support
Admission EventsRecord ValidatingAdmissionPolicy violation events received via the Kubernetes Audit Webhook
AlertsConfigure webhook alert rules to push security events to Slack (Discord can receive them through its Slack-compatible endpoint)
SyslogForward security events to an external syslog server via UDP or TCP
Event SourcesLive health of the Tetragon per-node streams and Hubble Relay, so "monitoring" never silently means "blind"; also lists whether Trivy Operator keeps producing reports
User ManagementUser account creation, role assignment, and JWT authentication management, with forced first-login password change and login rate limiting
Audit LogA record of every operation performed through K8s Sentinel (who, when, target, outcome), filterable and exportable to CSV
Event RetentionConfigure max event counts and TTL to control database storage usage

Target Audience​

  • DevSecOps Engineers: Personnel who need to rapidly define and adjust eBPF security policies in Kubernetes environments and continuously monitor security events
  • Platform Teams: Engineers responsible for operating Kubernetes clusters, managing Cilium network policies, and ensuring Tetragon security observability is functioning correctly
  • Teams adopting Tetragon in K8s: Technical staff who want to deploy Tetragon TracingPolicy in production but are unfamiliar with CRD operations or want to improve policy management efficiency

Reading Guide​

We recommend reading this documentation in the following order for the fastest path to deploying and using K8s Sentinel: