---
title: "Configure admission fair sharing in Red Hat build of Kueue 1.4"
description: "Red Hat build of Kueue 1.4 provides admission fair sharing for shared CPU, GPU, and high-performance computing clusters where FIFO admission can starve later workloads. The guide configures the feature through the Kueue custom resource, using either default values—30-minute usage half-life, five-minute sampling interval, and resource weights of 1—or custom settings, then creates a ResourceFlavor, a ClusterQueue using UsageBasedAdmissionFairSharing, and two LocalQueues. Kueue tracks each LocalQueue's historical consumption, adds a temporary penalty when a workload is admitted, and compares usage adjusted by queue weight to choose the lowest-use pending workload when quota returns. In the example, lq-a receives the first job, lq-b is admitted after job1 is deleted, and usage decays or grows on sampling ticks; usage ordering takes precedence over priority classes, which only break ties."
---

# Configure admission fair sharing in Red Hat build of Kueue 1.4

[Red Hat](https://yomu.fyi/company/red-hat) · Maysa De Macedo Souza · Sep 4, 2026

**Type:** Tutorial

## Summary

Red Hat build of Kueue 1.4 provides admission fair sharing for shared CPU, GPU, and high-performance computing clusters where FIFO admission can starve later workloads. The guide configures the feature through the Kueue custom resource, using either default values—30-minute usage half-life, five-minute sampling interval, and resource weights of 1—or custom settings, then creates a ResourceFlavor, a ClusterQueue using UsageBasedAdmissionFairSharing, and two LocalQueues. Kueue tracks each LocalQueue's historical consumption, adds a temporary penalty when a workload is admitted, and compares usage adjusted by queue weight to choose the lowest-use pending workload when quota returns. In the example, lq-a receives the first job, lq-b is admitted after job1 is deleted, and usage decays or grows on sampling ticks; usage ordering takes precedence over priority classes, which only break ties.

## Context

Shared clusters with constrained GPUs or high-spec CPUs can have multiple tenants competing for limited quota. The source states that FIFO admission may cause long-term starvation for workloads submitted later and reduce service predictability.

## Approach / What changed

Enable admissionFairSharing in the Kueue custom resource with Default or Custom settings, create a uniform ResourceFlavor and a ClusterQueue using UsageBasedAdmissionFairSharing, then place workloads in LocalQueues sharing that ClusterQueue. Kueue uses historical LocalQueue consumption, temporary admission penalties, sampling, half-life decay, and optional weights to order pending workloads.

## Takeaways

- Admission fair sharing reorders workloads at admission time without evicting running workloads; fair sharing-based preemption can evict them.
- The default configuration uses a 30-minute usage half-life, a five-minute sampling interval, and a resource weight of 1 for each resource; these defaults may change.
- Usage-based ordering across LocalQueues takes precedence over priority classes, with priority used as a tiebreaker when usage is equal.

**Tags:** [Kubernetes](https://yomu.fyi/topic/kubernetes), [OpenShift](https://yomu.fyi/topic/openshift)

- Source: [Red Hat](https://developers.redhat.com/articles/2026/09/04/configure-admission-fair-sharing-red-hat-build-kueue-1-4)
- Source URL: https://developers.redhat.com/articles/2026/09/04/configure-admission-fair-sharing-red-hat-build-kueue-1-4
- Ingested by Yomu: 2026-09-04T08:00:50.438Z

[Read original post](https://developers.redhat.com/articles/2026/09/04/configure-admission-fair-sharing-red-hat-build-kueue-1-4)