Behind the queues: How Kueue reimagines scheduling in Red Hat OpenShift

Behind the queues: How Kueue reimagines scheduling in Red Hat OpenShift Topology-aware scheduling Kueue with dynamic resource allocation Why it matters Next steps Red Hat OpenShift Container Platform | Product Trial About the authors Pannaga Rao Bhoja Ramamanohara Sohan Kunkerkar More like this NGINX Gateway Fabric certified for Red Hat OpenShift DxOperator from DH2i is now certified for Red Hat OpenShift 4.19 Where Coders Code | Command Line Heroes What Kind of Coder Will You Become? | Command Line Heroes Keep exploring Browse by channel Automation Artificial intelligence Open hybrid cloud Security Edge computing Infrastructure Applications Virtualization Share In a modern cluster, the hardest problem isn’t running workloads—it's sharing resources fairly. Red Hat OpenShift clusters are seeing a surge of AI-accelerated workloads, from GPU-intensive training jobs to large batches of inference requests. At the same time, other tenants still need consistent throughput for their everyday CI/CD pipelines and data processing tasks. The result is a constant battle for resources, where some jobs wait too long, others consume more than their fair share, and administrators are left fighting bottlenecks. This is exactly the challenge that Kueue, a Kubernetes-native job queueing and scheduling framework, was built to solve. It introduces structured queues, priorities, and quota enforcement to bring fairness and predictability back into scheduling. With Red Hat Build of Kueue, these upstream innovations are packaged, hardened, and delivered into Red Hat OpenShift as a supported, enterprise-ready solution to enable clusters to run efficiently while giving every workload a fair chance. Once workloads are queued fairly, the next challenge is where they actually land. For distributed jobs, placement can matter as much as allocation: pods that constantly exchange data perform very differently depending on whether they're co-located or scattered across zones. This is where topology-aware scheduling (TAS) comes in. Rather than treating the cluster as a flat pool of machines, TAS considers the physical and logical layout of the infrastructure (racks, blocks, zones) and makes scheduling decisions that optimize communication and efficiency. Workloads that talk a lot can be placed closer together, multi-pod jobs can start in sync through gang scheduling, and fairness across tenants is preserved even as locality is optimized.