[ANNOUNCE] kueue 0.18.0 is released

[Michał Woźniak]

We have just released Kueue v0.18.0.

For more details, please check the release notes and the documentation.

Highlights:

1. Topology-Aware Scheduling: Continued TAS maturation with preferred node-affinity support, better LeaderWorkerSet placement, and support for overlapping ResourceFlavors. Performance of snapshotting is significantly improved.
2. Scheduling Performance: SchedulingEquivalenceHashing graduates to beta and is enabled by default for batch-handling of scheduling-equivalent workloads. This is a critical performance improvement for large-scale deployments, where the Kueue scheduler previously spent significant time evaluating equivalent workloads.
3. DRA and Accelerators: The integration graduates to beta. Kueue also adds alpha support for DRA extended resources and partitionable devices with counter-based quota.
4. MultiKueue: MultiKueue improves reliability and automation with quota aggregation based on worker clusters, better reconnect handling, and more robust eviction.
5. Elastic Workloads: ElasticJobsViaWorkloadSlices graduates to beta, with significant hardening of KubeRay support.
6. Concurrent Admission: introduces alpha support for migrating admitted workloads between ResourceFlavors to chase the optimal available capacity. This can help reduce costs by moving already-running workloads from autoscaled nodes to reserved capacity during off-peak hours.

7.  Observability and UX: New metrics cover workload creation latency, pending resources, eviction latency, and queue/cohort information. KueueViz receives stability and usability fixes.

8. Admission Controls: New alpha-level admission controls provide more flexibility for workload admission. `quotaCheckStrategy` can restrict admission checks to resources explicitly declared in the ClusterQueue specification, while `WorkloadPriorityClassDefaulting` makes the WorkloadPriorityClass named `default` the default for Kueue-managed workloads.

9. Priority Booster Controller: Kueue publishes an example controller demonstrating how custom controllers can manipulate Kueue Workload priorities. The example shows how to achieve time sharing within a ClusterQueue.

Thank you to the contributors who reported issues, implemented features, fixed bugs, etc.

Cheers,
Michal