Kueue

Kueue is a set of APIs and controller for job queueing. It is a job-level manager that decides when a job should be admitted to start (as in pods can be created) and when it should stop (as in active pods should be deleted).

Read the overview to learn more.

Production Readiness status

• ✔️ Coverage with unit and integration tests: ~72%

• ✔️ Up-to-date documentation.

• ✔️ Monitoring via metrics

• ✔️ API validation and defaulting.

• ⏳ Scalability tests in progress.

• ✔️ API version: v1beta1, with backwards-compatibility guarantees.

  Based on community feedback, we continue to simplify and evolve the API to address new use cases.

Installation

Requires Kubernetes 1.22 or newer.

To install the latest release of Kueue in your cluster, run the following command:

kubectl apply -f https://github.com/kubernetes-sigs/kueue/releases/download/v0.2.1/manifests.yaml

The controller runs in the kueue-system namespace.

Read the installation guide to learn more.

Usage

A minimal configuration can be set by running the samples:

kubectl apply -f config/samples/single-clusterqueue-setup.yaml

Then you can run a job with:

kubectl create -f config/samples/sample-job.yaml

Learn more about:

• Kueue concepts.
• Common and advanced tasks.

Architecture

Learn more about the architecture of Kueue with the following design docs:

• bit.ly/kueue-apis discusses the API proposal and a high level description of how Kueue operates. Join the mailing list to get document access.
• bit.ly/kueue-controller-design presents the detailed design of the controller.

Roadmap

This is a high-level overview of the main priorities for 2023, in expected order of release:

• Job preemption to reclaim borrowed quota and to accommodate high priority jobs #83, this is planned for v0.3
• Cooperative preemption support for workloads that implement checkpointing #477
• Flavor assignment strategies, e.g. minimizing cost vs minimizing borrowing #312
• Integration with cluster-autoscaler for guaranteed resource provisioning
• Integration with common custom workloads #74:
  • Kubeflow (TFJob, MPIJob, etc.)
  • Spark
  • Ray
  • Workflows (Tekton, Argo, etc.)

These are features that we aim to have in the long-term, in no particular order:

• Budget support #28
• Dashboard for management and monitoring for administrators
• Multi-cluster support

Community, discussion, contribution, and support

Learn how to engage with the Kubernetes community on the community page.

You can reach the maintainers of this project at:

• Slack
• Mailing List

Code of conduct

Participation in the Kubernetes community is governed by the Kubernetes Code of Conduct.