Repository Issue Activity (beta)

#1755 feat(scheduler): checkpoint scenario-generator progress across scheduling sessions

#2120 Feasibility-aware tiebreak for subgroup K-cutoff selection

#2122 Extended resources are scaled by 1000 in MaxNodePoolResources validation

#1585 Add new Semi-Preemptible 3rd preemptibility mode

#1865 Investigate scale peformence degradation

#904 Expected Runtime Plugin for Soft Eviction via Requeue Action

#2065 [design] design issue to allow reclaim/consolidation from lower priority queues (even within quota)

#1186 Track flaky tests

#2118 Flaky env-test: Binder shutdown times out after EnvTests complete

#1772 Add scale tests for 1k nodes

#2100 DRA structured allocator hardcodes empty Features{}, breaking sharedCounters/PartitionableDevices allocation

#1603 Research Numa topology solution

#2115 scheduler: Releasing pods stay in the inter-pod affinity index, so reclaim/preempt can never satisfy a required podAntiAffinity against the victims

#2109 Allow deployments to be grouped under a single podGroup

#2077 Add topology-aware-scheduling - non-indexed block topology/automatic sub-grouping

#2098 Allow declaring gang size (minMember) on workloads handled by DefaultGrouper, e.g. StatefulSet

#1658 Cross ResourceClaim device topology constraints within a PodGroup

#1516 Skill to investigate a job's pending status

#1872 Make deferred resize a scheduling event KAI considers when pre-empting and reclaiming

#2020 Design of 'Gang of gangs' pattern where multiple workload types should be scheduled as a gang

#1760 feat(scheduler): add TopologyFirst scenario generator

#2106 cannot install helm chart with fips mode only

#2073 Separating consolidation from priority preemption

#2054 [Design] Queue Soft Limit - solving multi level SLAs and lower-priority queue over-quota starvation

#1933 How to schedule a pod onto a cordoned node?