WhatBrokeToday
GTA 6FAQ
Kubernetes

Kubernetes DynamicResources Plugin Assigns Wrong ResourceClaims to Pods

medium

Is it fixed? Not yet.

Ongoing since Nov 12, 09:44 UTC. First reported Nov 12, 09:44 UTC.

What happened

A data-correctness bug was identified in the DynamicResources scheduler plugin when a pod mixes a pending shared allocation with a new claim under specific feature gates. This causes the scheduler to assign allocation results to the wrong ResourceClaims.

What caused it?

Under active investigation. No report has identified a root cause yet.

No verified workaround yet

None of the linked reports describe a mitigation. This page updates when one appears.

Timeline

  1. Investigating(current)
  2. Workaround available(upcoming)
  3. Resolved(upcoming)
  1. First report on GitHub issue

    Reserve assigns allocation results to the wrong ResourceClaims when a pod mixes a pending shared allocation with a new claim

    Source

Evidence

1 report · 1 GitHubShow
  1. [1]
    GitHub First report
    Reserve assigns allocation results to the wrong ResourceClaims when a pod mixes a pending shared allocation with a new claim(opens GitHub issue in a new tab)

    /kind bug /sig scheduling /sig node /wg device-management What happened: While studying how the scheduler handles a ResourceClaim shared by two pods in a PodGroup, I hit what looks like a data-correctness bug in the DynamicResources plugin (feature gates DRAWorkloadResourceClaims=true + GenericWorkload=true, scheduling.k8s.io/v1alpha2 served). My test cluster runs two simulated GPUs (gpu-0, gpu-1) published by the NVIDIA DRA driver on top of nvml-mock (mokka). Two claims, each pinned to…

    by zl86790 0upvotes 3comments

FAQ

What happened in the Kubernetes issue?

A data-correctness bug was identified in the DynamicResources scheduler plugin when a pod mixes a pending shared allocation with a new claim under specific feature gates. This causes the scheduler to assign allocation results to the wrong ResourceClaims.

What is affected by the Kubernetes issue?

Kubernetes.

When did the Kubernetes issue start and end?

It started Nov 12, 09:44 UTC and is still ongoing (6 d 18 h so far). This page updates as it develops.

How do I get alerts for Kubernetes?

Subscribe to the Kubernetes RSS feed for every new Kubernetes incident, or connect an AI assistant to WhatBrokeToday over MCP and ask whether Kubernetes is down. Kubernetes feed

Where does this information come from?

WhatBrokeToday collects public status pages, developer community reports and security advisories, groups the ones about the same problem, and links every report under Evidence on this page. This one has 1 report from 1 independent source.