|
| 1 | +## v0.14.8 |
| 2 | + |
| 3 | +Changes since `v0.14.7`: |
| 4 | + |
| 5 | +## Changes by Kind |
| 6 | + |
| 7 | +### Deprecation |
| 8 | + |
| 9 | +- LWS: disable testing the mutating of the `kueue.x-k8s.io/workloadpriorityclass` label as the functionality is broken on |
| 10 | + Kueue 0.14 with Kubernetes 1.35+. |
| 11 | + |
| 12 | + If you are using this functionality, please migrate to use Kueue 0.15+. (#8541, @mimowo) |
| 13 | + |
| 14 | +### Feature |
| 15 | + |
| 16 | +- CLI: Support "kwl" and "kueueworkload" as a shortname for Kueue Workloads. (#8473, @kannon92) |
| 17 | + |
| 18 | +### Bug or Regression |
| 19 | + |
| 20 | +- Add lws editer and viewer roles to kustomize and helm (#8554, @kannon92) |
| 21 | +- Fix ClusterQueue deletion getting stuck when pending workloads are deleted after being assumed by the scheduler (#8552, @sohankunkerkar) |
| 22 | +- HC: Avoid redundant requeuing of inadmissible workloads when multiple ClusterQueues in the same cohort hierarchy are processed. (#8512, @sohankunkerkar) |
| 23 | +- Integrations based on Pods: skip using finalizers on the Pods created and managed by integrations. |
| 24 | + |
| 25 | + In particular we skip setting finalizers for Pods managed by the built in Serving Workloads Deployments, |
| 26 | + StatefulSets, and LeaderWorkerSets. |
| 27 | + |
| 28 | + This improves performance of suspending the workloads, and fixes occasional race conditions when a StatefulSet |
| 29 | + could get stuck when deactivating and re-activating in a short interval. (#8568, @mbobrovskyi) |
| 30 | +- JobFramework: Fixed a bug that allowed a deactivated workload to be activated. (#8445, @chengjoey) |
| 31 | +- Kubeflow TrainJob v2: fix the bug to prevent duplicate pod template overrides when starting the Job is retried. (#8488, @j-skiba) |
| 32 | +- LeaderWorkerSet: Fixed a bug that prevented deleting the workload when the LeaderWorkerSet was scaled down. (#8672, @mbobrovskyi) |
| 33 | +- MultiKueue now waits for WorkloadAdmitted (instead of QuotaReserved) before deleting workloads from non-selected worker clusters. To revert to the previous behavior, disable the `MultiKueueWaitForWorkloadAdmitted` feature gate. (#8601, @IrvingMg) |
| 34 | +- MultiKueue: fix the eviction when initiated by the manager cluster (due to eg. Preemption or WairForPodsReady timeout). (#8403, @mbobrovskyi) |
| 35 | +- ProvisioningRequest: Fixed a bug that prevented events from being updated when the AdmissionCheck state changed. (#8405, @mbobrovskyi) |
| 36 | +- TAS: Fix a bug that MPIJob with runLauncherAsWorker Pod indexes are not correctly evaluated during rank-based ordering assignments. (#8662, @tenzen-y) |
| 37 | +- TAS: Fixed an issue where workloads could remain in the second-pass scheduling queue (used for integration |
| 38 | + or TAS with ProvisioningRequests, and for TAS Node Hot Swap) even if they no longer require to be in the queue. (#8431, @skools-here) |
| 39 | +- TAS: fix TAS resource flavor controller to extract only scheduling-relevant node updates to prevent unnecessary reconciliation. (#8454, @Ladicle) |
| 40 | +- TAS: significantly improves scheduling performance by replacing Pod listing with an event-driven |
| 41 | + cache for non-TAS Pods, thereby avoiding expensive DeepCopy operations during each scheduling cycle. (#8484, @gabesaba) |
| 42 | + |
1 | 43 | ## v0.14.7 |
2 | 44 |
|
3 | 45 | Changes since `v0.14.6`: |
|
0 commit comments