WhatBrokeToday
GTA 6FAQ
Kubernetes

Kubernetes 1.37 DRA Pods Wait ~5 Minutes for Devices from Deleted Pods

medium

Is it fixed? Not yet.

Ongoing since Nov 13, 06:59 UTC. First reported Nov 13, 06:59 UTC.

What happened

In Kubernetes 1.37, when a Pod waiting for BindingConditions in PreBind() is deleted, a newly created Pod requesting the same device remains Pending for approximately 5 minutes even though the device is freed within seconds. This regression affects Dynamic Resource Allocation (DRA) workflows.

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

    DRA: Pod waits ~5 min for a device released by a deleted Pod that was waiting on BindingConditions [1.37 regression]

    Source

Evidence

1 report · 1 GitHubShow
  1. [1]
    GitHub First report
    DRA: Pod waits ~5 min for a device released by a deleted Pod that was waiting on BindingConditions [1.37 regression](opens GitHub issue in a new tab)

    What happened? When a Pod that is waiting for BindingConditions in PreBind() is deleted, another Pod that needs the same device stays Pending for about 5 minutes, although the actual device becomes free within seconds. Example case (see the reproduction section) The cluster has one device, whose BindingConditions are never satisfied. One Pod (a) is allocated the device and waits in PreBind(). I deleted a, and created another Pod (b) that requests the same device right away. On v1.37.0, the…

    by logica0419 0upvotes 7comments

FAQ

What happened in the Kubernetes issue?

In Kubernetes 1.37, when a Pod waiting for BindingConditions in PreBind() is deleted, a newly created Pod requesting the same device remains Pending for approximately 5 minutes even though the device is freed within seconds. This regression affects Dynamic Resource Allocation (DRA) workflows.

What is affected by the Kubernetes issue?

Kubernetes (versions: v1.37.0).

When did the Kubernetes issue start and end?

It started Nov 13, 06:59 UTC and is still ongoing (5 d 21 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.