AWS Parallel Computing Service enhanced its Slurm cluster management by supporting in-place upgrades up to three major Slurm versions. This process updates controllers, databases, and APIs while allowing existing compute jobs to continue uninterrupted and maintaining accounting data integrity.
Users can upgrade cluster control planes through the console, CLI, or API, then update compute nodes on their own schedule. This feature is designed to minimize disruption in high performance computing environments, facilitating easier cluster maintenance across AWS regions supporting PCS.
Source & further reading
AWS Parallel Computing Service supports in-place Slurm major version upgrades ↗Source published 2026-06-30. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.