Skip to main content

Release Announcement v1.7.0

We are pleased to announce that the Apache YuniKorn community has voted to release 1.7.0. Apache YuniKorn is a standalone resource scheduler, designed for managing and scheduling Batch and Data workloads on container orchestration frameworks like Kubernetes for on-prem and on-cloud use cases.

Overview

The Apache YuniKorn community has resolved 168 JIRAs in this release.

Release manager: Peter Bacsko

Release date: 2025-07-28

Highlights

Kubernetes 1.32 support

Apache YuniKorn 1.7.0 add supports for Kubernetes 1.32. See YUNIKORN-2996 for details.

Latest in-place pod vertical scaling support

Modifying pod resource usage in-place is now supported according to the latest spec (alpha in v1.32). See YUNIKORN-3053 for details.

Pod level resources limits support (KEP-2837)

With Kubernetes 1.32, it's possible to set resource limits on a pod level instead of a container level. See YUNIKORN-3039 for details.

GOGC / GOMEMLIMIT can be adjusted in the helm chart

Two environment variables, GOGC and GOMEMLIMIT are now easy to modify in the helm chart. See YUNIKORN-3000 for details.

Proper tracking of non-Yunikorn (foreign) pods

Before 1.7.0, only the total resource usage of foreign pods has been tracked. With the new release, Apache YuniKorn properly tracks non-Yunikorn pods in the scheduler core, enhancing supportability. Pods scheduled by different schedulers are listed in the REST response when doing state dump or listing nodes. See YUNIKORN-2791 for details.

Metrics improvements

Metrics for scheduling cycle and node evaluation was added in YUNIKORN-2930. Metric for maxRunningApps was added in YUNIKORN-2854.