JobSet

JobSet is a Kubernetes-native API for managing a group of k8s Jobs as a unit. It aims to offer a unified API for deploying HPC (e.g., MPI) and AI/ML training workloads (PyTorch, Jax, Tensorflow etc.) on Kubernetes.

Take a look at the concepts page for a brief description of how to use JobSet.

Conceptual Diagram

Features overview

• Support for multi-template jobs: JobSet models a distributed training workload as a group of K8s Jobs. This allows a user to easily specify different pod templates for different distinct groups of pods (e.g. a leader, workers, parameter servers, etc.), something which cannot be done by a single Job.

• Automatic headless service configuration and lifecycle management: ML and HPC frameworks require a stable network endpoint for each worker in the distributed workload, and since pod IPs are dynamically assigned and can change between restarts, stable pod hostnames are required for distributed training on k8s, By default, JobSet uses IndexedJobs to establish stable pod hostnames, and does automatic configuration and lifecycle management of the headless service to trigger DNS record creations and establish network connectivity via pod hostnames.

• Configurable success policies: JobSet has configurable success policies which target specific ReplicatedJobs, with operators to target Any or All of their child jobs. For example, you can configure the JobSet to be marked complete if and only if all pods that are part of the “worker” ReplicatedJob are completed. This enables users to use their compute resources more efficiently, allowing a workload to be declared successful and release the resources for the next workload more quickly.

• Configurable failure policies: JobSet has configurable failure policies which allow the user to specify a maximum number of times the JobSet should be restarted in the event of a failure. If any job is marked failed, the entire JobSet will be recreated, allowing the workload to resume from the last checkpoint. When no failure policy is specified, if any job fails, the JobSet simply fails.

• Exclusive Placement Per Topology Domain: JobSet includes an annotation which can be set by the user, specifying that there should be a 1:1 mapping between child job and a particular topology domain, such as a datacenter rack or zone. This means that all the pods belonging to a child job will be colocated in the same topology domain, while pods from other jobs will not be allowed to run within this domain. This gives the child job exclusive access to computer resources in this domain. You can run this example yourself to see how exclusive placement works.

• Fast failure recovery: JobSet recovers from failures by recreating all the child Jobs. When scheduling constraints such as exclusive Job placement are used, fast failure recovery at scale can become challenging. As of JobSet v0.3.0, JobSet uses a designed such that it minimizes impact on scheduling throughput. We have benchmarked scheduling throughput during failure recovery at 290 pods/second at a 15k node scale.

• Startup Sequencing: As of JobSet v0.5.0 users can configure a startup order for the ReplicatedJobs in a JobSet. This enables support for patterns like the “leader-worker” paradigm, where the leader must be running before the workers should start up and connect to it.

• Integration with Kueue: Use JobSet v0.2.3+ and Kueue v0.5.0+ to oversubscribe your cluster with JobSet workloads, placing them in queue which supports multi-tenancy, resource sharing and more. See Kueue documentation for more details on the benefits of managing JobSet workloads via Kueue.

Production Readiness status

• ✔️ API version: v1alpha2, respecting Kubernetes Deprecation Policy
• ✔️ Maintains support for latest 3 Kubernetes minor versions.
• ✔️ Up-to-date documentation.
• ✔️ Test Coverage:
  • ✔️ Unit Test testgrid.
  • ✔️ Integration Test testgrid
  • ✔️ E2E Tests for Kubernetes 1.27, 1.28 1.29 on Kind.
• ✔️ Monitoring via metrics.
• ✔️ Security: RBAC based accessibility.
• ✔️ Stable release cycle(2-3 months) for new features, bugfixes, cleanups.

Installation

Requires a Kubernetes cluster running one of the last 3 Kubernetes minor versions.

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

kubectl apply --server-side -f https://github.com/kubernetes-sigs/jobset/releases/download/v0.5.0/manifests.yaml

The controller runs in the jobset-system namespace.

Read the installation guide to learn more.

Roadmap

Below is the JobSet roadmap for 2024, in expected order of release:

• Support for MultiKueue (kubernetes-sigs#440)
• JobSet TTL to clean up completed workloads (kubernetes-sigs#279)
• Configurable Failure Policy (kubernetes-sigs#262)
• Graduate API to v1 (kubernetes-sigs#380)
• In-place restarts (kubernetes-sigs#467)
• Placement Policy API (kubernetes-sigs#75)
• Support for JobSet extensions (kubernetes-sigs#146)
  • PyTorch Extension
  • Jax Extension
  • MPI Extension

Troubleshooting common issues

See the troubleshooting guide for help resolving common issues.

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.