
Security
Kubernetes policy as a rule table, with the proof built in.
Firewall Policies, Universal Rules and Security Objects give the cluster a rule table your team can review. Impact Analysis shows what a change will deny before it goes live, and Assurance, Tetragon and compliance reports prove it keeps working.

Security in the console












Deep dive
Policy, from authoring to audit.
Firewall Policies
The rule table you know, with impact analysis before every publish.
Sections and rules you can reorder, enable and disable, with source, destination, service and applied-to, L7 and context profiles. A default policy per namespace, View YAML and Export to Excel.
- Tabs: "Network Policy (CNP/INP/OVN)" and "Clusterwide Policy"
- "Hit Count" and "Risks" on every rule
- "Publish Validation" → "Impact Analysis" ("Impacted Endpoints", "New Denied Flows") → Publish
- "Cluster baseline protection": essential allowances that heal drift from out-of-band deletes
- Private Network scoping on Isovalent; manage policy through a linked Git repository
Universal Rules · Security Objects
Write a rule once, publish it to every cluster, and catch drift.
Universal Rules publish one rule set to many clusters, with a preview per cluster. The building blocks are shared Security Objects, so a rule reads like a firewall rule, not a selector.
- Publish preview per cluster
- "Check Drift", a drift schedule and "Restore Rules"
- Labels, Services, DNS, L7, Security Groups and CIDRs
- On Isovalent: Pod Nets, FQDN Groups and External Endpoints
Topology · Observatory · Locator
From the traffic you can see to the policy you need.
Live Hubble flows on the topology become prefilled rules. Observatory records an app's flows and proposes the allow policy for you to review before applying.
- "Create Allow/Deny policy prefilled from this flow"
- Observatory: record Hubble flows, alerts, "Export to Canvas"
- "Policy Gap Analysis" proposes an allow policy, reviewed before it is applied
- Locator: which rules affect given traffic
Assurance · Tetragon · Compliance
See it, fix it, prove it.
Monitors define what good looks like, Violations catch drift with evidence, Tetragon shows what is actually running, and the compliance report turns it into audit-ready evidence.
- Monitors: workload identity, required and prohibited labels, image, port and security-context rules; "Baseline from workload"
- "Connectivity Monitor" and "Runtime Monitor" (Tetragon allowed processes), auto-remediation by label
- Violations: acknowledge, suppress, reopen, with evidence and a lifecycle timeline
- Workload detail (Tetragon): "Listening ports", "Recent processes", "Privilege events"
- Compliance report: Executive Summary score, Framework Mappings, Enforcement Evidence, Drift Analysis; Excel, JSON and PDF. Change Audit: old vs new object per change
Fine print
- Pod Nets, FQDN Groups, External Endpoints and Private Network scoping are Isovalent features.
- Assurance works on any Kubernetes datasource; Registry is for the Cilium family.
See it live