Feature reference

This page is generated from the Kwatch feature catalog. Feature IDs are stable product vocabulary and should not be renamed without a migration.

Feature ID	Lifecycle	Description	Dependencies
core.detection.pods	runtime	Detect pod and container failures	none
core.pods.scheduling	runtime	Detect pod scheduling failures	core.detection.pods
core.pods.oom	runtime	Detect out-of-memory kills and repeated OOMs	core.detection.pods
core.pods.readiness	runtime	Detect sustained pod readiness failures	core.detection.pods
core.pods.restarts	runtime	Detect crash loops and excessive restarts	core.detection.pods
core.detection.workloads	runtime	Detect workload rollout and execution failures	none
core.workloads.rollouts	runtime	Detect stuck workload rollouts	core.detection.workloads
core.workloads.jobs	runtime	Detect failed Jobs and failed or missed CronJobs	core.detection.workloads
core.workloads.pdb	runtime	Detect PodDisruptionBudgets that block disruption	core.detection.workloads
core.workloads.hpa	runtime	Diagnose HorizontalPodAutoscaler failures	core.detection.workloads
core.detection.nodes	runtime	Detect node readiness, pressure and draining	none
core.nodes.usage	runtime	Detect node resource pressure from kubelet stats	core.detection.nodes
core.detection.storage	runtime	Detect claim, volume and attachment failures	none
core.storage.usage	runtime	Detect volumes nearly full	core.detection.storage
core.detection.network	runtime	Detect Service, Ingress and NetworkPolicy failures	none
core.detection.admission	runtime	Detect admission webhook and policy failures	none
core.detection.certificates	runtime	Detect expiring and invalid TLS certificates	none
core.detection.quota	runtime	Detect exhausted ResourceQuotas and LimitRange rejections	none
core.detection.custom-resources	runtime	Detect failing conditions on custom and built-in resources	none
core.detection.control-plane	runtime	Check API server, etcd, DNS, scheduler and controller-manager	none
core.detection.probes	runtime	Run configured HTTP, TCP and DNS checks	none
core.detection.risks	runtime	Quote configuration risks (no probe, no limit, mutable tag, single replica) when a failure shows their cost	core.detection.workloads
core.detection.unusual-events	runtime	Report repeated Warning events kwatch has no detector for, quoting the event	none
analysis.root-cause	runtime	Explain each incident by its most likely root cause	none
analysis.shared-factors	runtime	Suspect a node that workloads failing together share when nothing else explains them	analysis.root-cause
analysis.changes	runtime	Relate incidents to recent changes and who made them	analysis.root-cause
analysis.impact	runtime	Show which workloads and services an incident affects	analysis.root-cause
analysis.investigation	runtime	Attach a redacted log excerpt to announcements	none
incidents.noise-control	runtime	Settle, merge symptoms and send only material changes	analysis.root-cause
incidents.flapping	runtime	Recognise flapping and routine recurring incidents	incidents.noise-control
incidents.startup-summary	startup	Summarise pre-existing incidents once at cold start	none
incidents.state	startup	Keep incidents and history on disk across restarts	none
incidents.downtime-changes	startup	Report changes made while kwatch was down	incidents.state
policy.scope	runtime	Filter by namespace, reason, selector and silences	none
policy.severity	runtime	Override severity by reason and owner kind	none
policy.maintenance	runtime	Hold incidents for objects under maintenance	none
policy.runbooks	runtime	Attach reason-aware runbook links	none
delivery.routing	runtime	Route incidents to providers by scope	none
delivery.threads	runtime	Update one message thread per incident where supported	none
delivery.templates	runtime	Render operator-selected text templates	none
delivery.audit-log	runtime	Write a JSON line for every incident decision	none
security.rbac-audit	runtime	Report missing permissions	none