koordinator-sh/koordinator

A QoS-based scheduling system brings optimal layout and status to workloads such as microservices, web services, big data jobs, AI jobs, etc.

What it solves

Koordinator addresses the challenge of managing hybrid orchestration workloads on Kubernetes, specifically the need to balance latency-sensitive services and batch jobs on the same cluster. It aims to reduce resource waste, minimize interference between containers, and simplify the complex manual tuning of resource configurations.

How it works

It functions as a QoS-based scheduling system that enhances Kubernetes by implementing advanced algorithms and flexible scheduling policies. This allows the system to optimize how pods are deployed and increase deployment density, ensuring that cluster resources are used more effectively while maintaining the reliability of high-priority workloads.

Who it’s for

It is designed for Kubernetes administrators and DevOps engineers who need to run mixed workloads (such as Spark jobs alongside web services) and want to improve resource utilization and system performance without extensive manual tuning.

Highlights

  • QoS-based scheduling: Optimizes the placement of hybrid workloads to ensure reliability.
  • Improved resource utilization: Increases pod deployment density to make better use of available hardware.
  • Reduced interference: Uses techniques to prevent batch jobs from impacting the performance of latency-sensitive applications.
  • Flexible policies: Provides customizable scheduling options to suit specific organizational needs.
  • Easy integration: Designed to be added to existing Kubernetes clusters with minimal friction.

Related

  • Project
    kai-scheduler/KAI-Scheduler

    A scalable Kubernetes scheduler that optimizes GPU resource allocation and fairness for large-scale AI and machine learning workloads.

  • Project
    volcano-sh/volcano

    A Kubernetes-native batch scheduling system that optimizes resource management for AI, machine learning, and Big Data workloads.

  • Project
    nebius/soperator

    A Kubernetes operator that automates the deployment and management of Slurm clusters for AI training and HPC, providing high reliability and efficient GPU utilization.

  • Project
    kubernetes-sigs/jobset

    A Kubernetes-native API for managing groups of Jobs as a single unit, specifically designed to simplify the deployment of distributed AI/ML training and HPC workloads.

  • Project
    fluid-cloudnative/fluid

    A Kubernetes-native distributed dataset orchestrator and accelerator that optimizes data access for AI and big data applications by providing dataset abstraction and elastic caching.