Tasks

The following tasks show you how to perform operations based on the Kueue user personas such as batch administrators and batch users.

Batch administrator

A batch administrator manages the cluster infrastructure and establishes quotas and queues.

As a batch administrator, you can learn how to:

• Setup role-based access control to Kueue objects.
• Administer cluster quotas with ClusterQueues and LocalQueues.
• Setup All-or-nothing with ready Pods.
• As a batch administrator, you can learn how to monitor pending workloads.
• As a batch administrator, you can learn how to run a Kueue managed Jobs with a custom WorkloadPriority.
• As a batch administrator, you can learn how to setup a MultiKueue environment.

Batch user

A batch user runs workloads. A typical batch user is a researcher, AI/ML engineer, data scientist, among others.

As a batch user, you can learn how to:

• Run a Kueue managed batch/Job.
• Run a Kueue managed Flux MiniCluster.
• Run a Kueue managed Kubeflow Job. Kueue supports MPIJob v2beta1, PyTorchJob, TFJob, XGBoostJob, PaddleJob, and MXJob.
• Run a Kueue managed KubeRay RayJob.
• Run a Kueue managed KubeRay RayCluster.
• Submit Kueue jobs from Python.
• Run a Kueue managed plain Pod.
• Run a Kueue managed JobSet.

Platform developer

A platform developer integrates Kueue with other software and/or contributes to Kueue.

As a platform developer, you can learn how to:

• Integrate a custom Job with Kueue.
• Enable pprof endpoints.

Troubleshooting

Sometimes things go wrong. You can follow the Troubleshooting guides to understand the state of the system.