Ross ROSS = Recommend OSS · open-source software intelligence for agents

volcano-sh/volcano

A Cloud Native Batch System (Project under CNCF) observed · 2026-08-28

github.com/volcano-sh/volcano · homepage · Go · Apache-2.0 (permissive) observed · 2026-08-28

Health v2 · maintenance only

98/100

  • Activity 99
  • Release rhythm 95
  • Longevity 100
How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-02. Adoption (stars, forks) is never an input.

  • gap_med: 22.0
  • age_days: 2729
  • days_rel: 34
  • days_push: 7
  • n_releases_24m: 25

Full methodology

Adoption not part of the score

5899 stars · 1506 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Volcano is a CNCF-hosted, Kubernetes-native batch scheduling system that extends kube-scheduler for high-performance workloads like AI/ML training, big data, and HPC. It provides gang scheduling, queue management, heterogeneous GPU/NPU scheduling, and multi-cluster scheduling with integrations for frameworks such as TensorFlow, PyTorch, Spark, Flink, and Ray.

Use cases

  • run distributed PyTorch or TensorFlow training jobs on Kubernetes with gang scheduling
  • schedule Spark and Flink batch workloads on a Kubernetes cluster
  • manage GPU sharing and heterogeneous device scheduling for AI workloads
  • implement multi-level queue and resource quota management for batch jobs
  • colocate online and offline workloads to improve cluster utilization
  • schedule jobs across multiple Kubernetes clusters
  • run HPC and genomics workloads on Kubernetes

When to choose

  • you need gang scheduling so all pods of a distributed job start together
  • you run AI/ML training, big data, or HPC batch workloads on Kubernetes
  • you need fine-grained queue management, quotas, and preemption for batch jobs
  • you need GPU/NPU scheduling with CUDA, MIG, or device sharing
  • you want to colocate online services and offline batch jobs

When to avoid

  • you only run simple microservices with default kube-scheduler needs
  • you are not using Kubernetes as your orchestration platform
  • you need a lightweight single-node scheduler without cluster infrastructure

Facets

service · maturity stable

container-orchestration scheduling cloud gpu-computing workflow-automation microservices cloud-computing machine-learning big-data go cloud batch-scheduling gang-scheduling hpc kubernetes-scheduler cncf gpu-scheduling queue-management distributed-training devops kubernetes linux docker

10 sources

Member repositories

RepositoryRoleHealth v2
volcano-sh/volcanomain98

For agents

markdown · JSON · MCP: product_card(name="volcano-sh/volcano")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem