Kubernetes: Scheduler Pod Tracking Fix
A critical scheduler bug fix addresses memory leaks in pod tracking when pods are recreated during scheduling failures. The fix ensures proper cleanup of in-flight pod references by tracking UIDs consistently.
Duration: PT1M47S
Episode overview
This episode is a short developer briefing from Kubernetes.
It explains recent repository work in plain language.
- Show: Kubernetes
- Published: 2026-04-12T00:00:00Z
- Audio duration: PT1M47S
Transcript excerpt
This excerpt keeps the crawler page concise. Listen to the episode or use the RSS feed for the full update.
Good morning, this is your Kubernetes development briefing for April 12th, 2026.
Today we have one significant commit from Samarth Verma addressing a critical scheduler issue. The fix resolves an inFlightPods memory leak that occurs when a pod is recreated during scheduling failure.
The problem centered around the scheduler's handling of pod UIDs during the scheduling process. When handleSchedulingFailure refreshes pod information from the informer before calling AddUnschedulableIfNotPresent, a pod could be deleted and recreated with the same name but a different UID. However, the inFlightPods…
The solution introduces an explicit in-flight UID parameter that's threaded through queueing-hint lookups, ensuring consistent tracking throughout the scheduling lifecycle. The commit includes comprehensive regression tests covering the same-name recreation scenario and updates error checking across affected test…
This fix touches seven files across the scheduler backend queue, scheduling logic, and integration tests. The changes include modifications to active_queue.go and scheduling_queue.go for the core logic, expanded test coverage in scheduling_queue_test.go with 78 new lines, and…
What's…
Nearby episodes from Kubernetes
- Scheduler Improvements and Resource Management Updates
- Testing Infrastructure and Scheduler Performance Updates
- etcd Upgrade and Test Cleanup
- Weekly Recap - Storage Stability & Scheduler Improvements
- Logging Level Adjustment for API Server
- Storage Version Fixes and Performance Improvements
- Weekly Recap - Performance Optimizations & Infrastructure Updates
- Under the Hood Improvements