50 Pull requests merged by 14 people

• Fix 'has malformed LC_DYSYMTAB' warnings for integration testings for MacOS

  #4523 merged Mar 8, 2025

• chore: Avoid empty slice usage in RaJob testings

  #4522 merged Mar 7, 2025

• chore: Avoid to initialize field.ErrorList with an empty slice

  #4526 merged Mar 7, 2025

• Create ResourceFlavor before Topology

  #4511 merged Mar 7, 2025

• Move Pod labels and annotation to separate package.

  #4472 merged Mar 7, 2025

• docs: create RemoveFinalizersWithStrictPatch KEP

  #4416 merged Mar 7, 2025

• Do not consider preemption.borrowWithinCohort in FairSharing preemptions

  #4165 merged Mar 7, 2025

• Record the full assignment usage to avoid recomputing

  #4510 merged Mar 7, 2025

• Fix integration test on release-0.10

  #4521 merged Mar 7, 2025

• Improve Scheduling and FairSharingIterator logging

  #4517 merged Mar 7, 2025

• Simply Clearing of FairSharingIterator's DRSValue map

  #4518 merged Mar 7, 2025

• Determine candidate ordering for Fair Sharing in Hierarchical Cohorts

  #4503 merged Mar 7, 2025

• Automated cherry pick of #4341: TAS: fix topology assignment for the RayJob's submitter Job #4498: Add check to verify RayJob tas pods were deleted

  #4502 merged Mar 6, 2025

• [Cleanup] Add listOptions to WaitForActivePodsAndTerminate

  #4513 merged Mar 6, 2025

• fix arguments to AgnHostImage in AppWrapper TAS test

  #4507 merged Mar 5, 2025

• TAS: support for cohorts and preemptions within them

  #4418 merged Mar 5, 2025

• add binding metrics port for helm and kustomize

  #4486 merged Mar 5, 2025

• [Flake] Extend VeryLongTimeout duration to 5 minutes

  #4504 merged Mar 5, 2025

• Add check to verify RayJob tas pods were deleted

  #4498 merged Mar 5, 2025

• Bump mui/material and mui/icons-material from 5.16.14 to 6.4.6 in /cmd/experimental/kueue-viz/frontend

  #4494 merged Mar 5, 2025

• Define EntryIterator

  #4493 merged Mar 5, 2025

• ClusterQueueSnapshot.DominantResourceShare - Delete unused return value

  #4496 merged Mar 5, 2025

• [release-0.9] fix construction of activePods in WaitForActivePodsAndTerminate

  #4492 merged Mar 5, 2025

• [release-0.10] fix construction of activePods in WaitForActivePodsAndTerminate

  #4491 merged Mar 5, 2025

• Automated cherry pick of #4438: [Flake] Change image behavior of high-priority-group pod

  #4452 merged Mar 5, 2025

• Automated cherry pick of #4484: Allow for one to change namespace for roles for prometheus

  #4488 merged Mar 5, 2025

• Automated cherry pick of #4478: [release-0.10] Use ReserveQuotaAt instead of ReserveQuota on unit tests

  #4490 merged Mar 5, 2025

• Automated cherry pick of #4484: Allow for one to change namespace for roles for prometheus

  #4487 merged Mar 5, 2025

• Allow for one to change namespace for roles for prometheus

  #4484 merged Mar 5, 2025

• [website] Doc: Fix Prometheus Operator installing guide link

  #4485 merged Mar 5, 2025

• Doc: Fix Prometheus Operator installing guide link

  #4483 merged Mar 5, 2025

• update jobset to 0.8 but force controller runtime to 0.19.4

  #4467 merged Mar 4, 2025

• fix construction of activePods in WaitForActivePodsAndTerminate

  #4482 merged Mar 4, 2025

• TAS: fix topology assignment for the RayJob's submitter Job

  #4341 merged Mar 4, 2025

• [release-0.10] Use ReserveQuotaAt instead of ReserveQuota on unit tests

  #4478 merged Mar 4, 2025

• Use ReserveQuotaAt instead of ReserveQuota on unit tests

  #4476 merged Mar 4, 2025

• Serialization support for FlavorResource, FlavorResourceSet and FlavorResourceQuantities

  #4420 merged Mar 4, 2025

• Fix flaky deploy ManageJobsWithoutQueue e2e test

  #4469 merged Mar 4, 2025

• Bump sigs.k8s.io/controller-tools from 0.17.1 to 0.17.2 in /hack/internal/tools

  #4463 merged Mar 3, 2025

• rename local variable in WaitForAppWrapperAvailability

  #4466 merged Mar 3, 2025

• Clean up TAS findLevelWithFitDomain loop

  #4464 merged Mar 3, 2025

• feat: add configuration to kueue-viz backend

  #4178 merged Mar 3, 2025

• Unexport manager fields

  #4460 merged Mar 3, 2025

• Bump github.com/golangci/golangci-lint from 1.63.4 to 1.64.6 in /hack/internal/tools

  #4461 merged Mar 3, 2025

• Refactor the accounting for preempted workloads in scheduler

  #4449 merged Mar 3, 2025

• Bump sigs.k8s.io/kind from 0.26.0 to 0.27.0 in /hack/internal/tools

  #4458 merged Mar 3, 2025

• Bump github.com/gohugoio/hugo from 0.144.2 to 0.145.0 in /hack/internal/tools

  #4456 merged Mar 3, 2025

• Introduce the helper struct for preemptedWorkloads

  #4447 merged Mar 3, 2025

• Automated cherry pick of #4438: [Flake] Change image behavior of high-priority-group pod

  #4450 merged Mar 3, 2025

• [Flake] Change image behavior of high-priority-group pod

  #4438 merged Mar 3, 2025

9 Pull requests opened by 7 people

• Differentiate error message for cohort and cluster queue

  #4473 opened Mar 4, 2025

• WIP TAS flake test

  #4475 opened Mar 4, 2025

• Update ResourceFlavor site documentation

  #4500 opened Mar 5, 2025

• ClusterQueueReference; CohortReference

  #4512 opened Mar 6, 2025

• [Cleanup] make terminating pods not racy

  #4515 opened Mar 6, 2025

• WIP: TAS: Use TAS implicitly when CQ is TAS-only

  #4519 opened Mar 7, 2025

• Add JobUID to preemption condition messages

  #4524 opened Mar 7, 2025

• self nominate for kannon92 for reviewer

  #4527 opened Mar 8, 2025

• use 8443 instead of target port

  #4528 opened Mar 8, 2025

13 Issues closed by 1 person

• Flaky Test: Topology Aware Scheduling when Delete Topology when ResourceFlavor exist should not allow to delete topology

  #4506 closed Mar 7, 2025

• Infinite preemption loop in fair sharing

  #3779 closed Mar 7, 2025

• Failing Test: MXJob controller when TopologyAwareScheduling enabled should admit workload which fits in a required topology domain

  #4344 closed Mar 7, 2025

• [kueuectl] Support filter by custom job type on the list pods command.

  #2806 closed Mar 7, 2025

• ClusterQueue not reclaiming resources from high-weight ClusterQueue in cohort

  #4333 closed Mar 6, 2025

• Flaky test: TopologyAwareScheduling for AppWrapper when Creating an AppWrapper Should place pods

  #4495 closed Mar 5, 2025

• Flaky test: TopologyAwareScheduling for RayJob when Creating a RayJob Should place pods based on the ranks-ordering

  #4497 closed Mar 5, 2025

• Flaky Test: Should run an appwrapper containing a job on worker if admitted

  #4477 closed Mar 4, 2025

• TAS: Implement e2e tests for RayJob with TAS

  #3716 closed Mar 4, 2025

• [Flaky unit test] TestPreemption/preempt_newer_workloads_with_the_same_priority

  #4425 closed Mar 4, 2025

• Problem serializing some Kueue types for logging

  #4137 closed Mar 4, 2025

• Flaky E2E Test: Should suspend the pods of a deployment created in the test namespace

  #4462 closed Mar 4, 2025

• Unexport access to Cohorts and ClusterQueues maps in hierarchy.Manager

  #2996 closed Mar 3, 2025

14 Issues opened by 7 people

• ProvisionRequest: GA Requirements

  #4529 opened Mar 8, 2025

• Flaky E2E Test: TopologyAwareScheduling when Creating a JobSet requesting TAS should admit a [Pod|PodGroup|JobSet] via TAS

  #4525 opened Mar 7, 2025

• Flaky test: StatefulSet created should allow to scale up after scale down to zero

  #4520 opened Mar 7, 2025

• Flaky Test (0.10): TopologyAwareScheduling for StatefulSet when Creating a StatefulSet Should place pods based on the ranks-ordering

  #4516 opened Mar 6, 2025

• Make WaitForActivePodsAndTerminate not racy

  #4514 opened Mar 6, 2025

• Increase the threshold for the output size for a failed ginkgo assert

  #4509 opened Mar 6, 2025

• Flaky Test: TopologyAwareScheduling for RayJob when Creating a RayJob Should place pods based on the ranks-ordering

  #4508 opened Mar 6, 2025

• [Flaky test] End To End Suite: kindest/node:v1.32.0: 4 failed e2e tests

  #4505 opened Mar 5, 2025

• Doc: Add Topology to Concept page

  #4501 opened Mar 5, 2025

• Update ResourceFlavor site documentation

  #4499 opened Mar 5, 2025

• Flaky e2e test: StatefulSet integration when StatefulSet created should admit group that fits

  #4480 opened Mar 4, 2025

• TAS: Add an algorithm that minimizes the fragmentation

  #4474 opened Mar 4, 2025

• WaitForPodsReady: support deleting workload when backoffLimitCount is reached

  #4471 opened Mar 4, 2025

• Upgrade controller-runtime to v0.20.x

  #4470 opened Mar 3, 2025

47 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• make local metrics configurable

  #4337 commented on Mar 7, 2025 • 26 new comments

• Optimise waiting for the availability of CRD-based frameworks

  #4059 commented on Mar 6, 2025 • 11 new comments

• TAS: allow to dump the usage of domains for TAS ResourceFlavors

  #4295 commented on Mar 8, 2025 • 5 new comments

• Add disabled `waitForPodsReady` to configuration by default.

  #4067 commented on Mar 3, 2025 • 0 new comments

• TAS: Better accuracy of scheduling by tighter integration with kube-scheduler

  #3755 commented on Mar 6, 2025 • 0 new comments

• [Flaky test] TopologyAwareScheduling when Creating a Pod requesting TAS should admit a single Pod via TAS

  #4140 commented on Mar 7, 2025 • 0 new comments

• v1beta2 wishlist

  #768 commented on Mar 7, 2025 • 0 new comments

• Admission Check controller to verify node capacity before admitting workloads

  #3211 commented on Mar 7, 2025 • 0 new comments

• Create an e2e customconfig test that tests WaitForPodsReady

  #4429 commented on Mar 7, 2025 • 0 new comments

• Add Support for Plain Pods with Known OwnerReferences

  #4106 commented on Mar 7, 2025 • 0 new comments

• Support KubeRay in MultiKueue via managedBy

  #3822 commented on Mar 7, 2025 • 0 new comments

• Make it easy to describe the preempting workload

  #4038 commented on Mar 7, 2025 • 0 new comments

• Revisit the documentation with respect to the concepts of Job and Workload

  #4286 commented on Mar 7, 2025 • 0 new comments

• MVP support for serving workloads running as LeaderWorkerSet

  #3232 commented on Mar 8, 2025 • 0 new comments

• ☂️ Release v0.11.0 features

  #4249 commented on Mar 8, 2025 • 0 new comments

• [KEP] Add ResourceFlavor fallbacks KEP

  #2561 commented on Mar 6, 2025 • 0 new comments

• Add RemoveFinalizersWithStrictPatch feature gate

  #3912 commented on Mar 8, 2025 • 0 new comments

• [kueue-viz] Change backend port to 8181

  #3926 commented on Mar 4, 2025 • 0 new comments

• Support custom scale-up and scale-down operations for StatefulSets.

  #3951 commented on Mar 8, 2025 • 0 new comments

• KEP-976: Support Groups of Pods Where Driver Generates Workers in Pod Integration

  #4102 commented on Mar 8, 2025 • 0 new comments

• add stand alone pod annotation check in pod webhook

  #4109 commented on Mar 8, 2025 • 0 new comments

• add ability to give tls certificates to metrics server

  #4385 commented on Mar 8, 2025 • 0 new comments

• KEP-4377: Secured Prometheus metrics endpoints

  #4404 commented on Mar 6, 2025 • 0 new comments

• upgrade autoscaling to v1

  #4444 commented on Mar 5, 2025 • 0 new comments

• Add integration test for visibility server

  #1377 commented on Mar 3, 2025 • 0 new comments

• Use ClusterQueueReference to improve static code analysis

  #4442 commented on Mar 3, 2025 • 0 new comments

• MultiKueue when Creating a multikueue admission check Should run an appwrapper containing a job on worker if admitted

  #4378 commented on Mar 3, 2025 • 0 new comments

• Support the use of Cert Manager

  #4164 commented on Mar 3, 2025 • 0 new comments

• Support JAX in Kueue using training-operator 1.9

  #4073 commented on Mar 4, 2025 • 0 new comments

• set nodeSelector for kueue-controller-manager via helm chart

  #3740 commented on Mar 4, 2025 • 0 new comments

• Support log retrieval on MultiKueue

  #3526 commented on Mar 4, 2025 • 0 new comments

• TAS: cache information is not updated when RF spec (nodeTaints or tolerations) is changed

  #3733 commented on Mar 4, 2025 • 0 new comments

• Enable sorting of queue items based on a custom timestamp extracted from an annotation

  #4415 commented on Mar 4, 2025 • 0 new comments

• Dashboard support in Kueue

  #940 commented on Mar 5, 2025 • 0 new comments

• Dynamic Job Parallelism and Resource Scaling Based on Backlog Metrics

  #2964 commented on Mar 5, 2025 • 0 new comments

• Python example of how to delete/stop kueue jobs that are queues or in progress

  #3875 commented on Mar 5, 2025 • 0 new comments

• Add configuration parameters to kueue-viz backend (listen-port, logs, verbosity, etc...)

  #3880 commented on Mar 5, 2025 • 0 new comments

• TAS: reduce friction by defaulting the PodSet annotations

  #3754 commented on Mar 6, 2025 • 0 new comments

• Flaky Test: Pod groups when Single CQ should allow to preempt the lower priority group

  #4434 commented on Mar 6, 2025 • 0 new comments

• FairSharing.Weight sensitive to high values

  #4247 commented on Mar 6, 2025 • 0 new comments

• reclaimWithinCohort preemption does not work with hierarchical cohorts

  #3948 commented on Mar 6, 2025 • 0 new comments

• [MultiKueue] Report a ClusterQueue as inactive (misconfigured) if there is ProvReq used with MK

  #2023 commented on Mar 6, 2025 • 0 new comments

• MultiKueue: Support sequential attempts to try worker clusters

  #3757 commented on Mar 6, 2025 • 0 new comments

• Support Hierarchical Cohorts with FairSharing

  #3759 commented on Mar 6, 2025 • 0 new comments

• Set `Requeued` condition, and unset `QuotaReserved` condition on eviction in Workload Reconciler

  #3474 commented on Mar 6, 2025 • 0 new comments

• Partial preemption of workloads

  #975 commented on Mar 6, 2025 • 0 new comments

• Serving-aware partial preemption of workloads

  #3762 commented on Mar 6, 2025 • 0 new comments