# Agent Kubernetes PodDisruptionBudgets and Evictions
Status: public
Confidence: medium (0.685) (verified)
Last verified: 2026-06-03
Generation: ai_structured


## TL;DR

PodDisruptionBudgets tell agents whether planned Kubernetes maintenance is allowed to evict pods without violating application availability rules.

## Core Explanation

Agents troubleshooting a stalled drain or a failed rollout need to distinguish application bugs from disruption policy. A healthy deployment can still block voluntary evictions when its PodDisruptionBudget says too few replicas would remain available.

The evidence set should include the PDB spec, selector match, current allowed disruptions, owning controller, desired replica count, pod readiness, eviction responses, and drain command output. Without that context, scaling or deleting pods can reduce availability or fight cluster policy.

## Source-Mapped Facts

- Kubernetes documentation distinguishes voluntary disruptions from involuntary disruptions when explaining pod availability. ([source](https://kubernetes.io/docs/concepts/workloads/pods/disruptions/))
- Kubernetes documentation says a PodDisruptionBudget can limit how many Pods of a replicated application are down simultaneously from voluntary disruptions. ([source](https://kubernetes.io/docs/concepts/workloads/pods/disruptions/))
- Kubernetes documentation says PodDisruptionBudgets use minAvailable or maxUnavailable to describe availability requirements. ([source](https://kubernetes.io/docs/tasks/run-application/configure-pdb/))

## Further Reading

- [Kubernetes Disruptions](https://kubernetes.io/docs/concepts/workloads/pods/disruptions/)
- [Kubernetes Configure a PodDisruptionBudget](https://kubernetes.io/docs/tasks/run-application/configure-pdb/)