fix(scheduler): Do not include resources with a count of 0.

CHANGELOG.md

@@ -5,6 +5,41 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [v0.9.13] - 2026-03-04
+## [Unreleased]
+
+### Fixed
+
+- Updated resource enumeration logic to exclude resources with count of 0. [#1120](https://github.com/NVIDIA/KAI-Scheduler/issues/1120)
+
+## [v0.13.0] - 2026-03-02
+### Added
+- Added `global.nodeSelector` propagation from Helm values to Config CR, ensuring operator-created sub-component deployments (admission, binder, scheduler, pod-grouper, etc.) receive the configured nodeSelector [#1102](https://github.com/NVIDIA/KAI-Scheduler/pull/1102) [yuanchen8911](https://github.com/yuanchen8911)
+- Added `plugins` and `actions` fields to SchedulingShard spec, allowing per-shard customization of scheduler plugin/action enablement, priority, and arguments [gshaibi](https://github.com/gshaibi)
+- Added support for Kubeflow Trainer v2 TrainJob workloads via skipTopOwner grouper pattern
+- Added `binder.cdiEnabled` Helm value to allow explicit override of CDI auto-detection for environments without ClusterPolicy
+- Added metric for tracking evicted pods in pod groups, including nodepool, eviction action, and gang size
+- Block scheduling of pods with shared (non-template) DRA GPU claims that lack a queue label or have a mismatched queue label [gshaibi](https://github.com/gshaibi)
+- Added the option to disable prometheus service monitor creation [#810](https://github.com/NVIDIA/KAI-Scheduler/pull/810) [itsomri](https://github.com/itsomri)
+- Fixed prometheus instance deprecation - ensure single instance [#779](https://github.com/NVIDIA/KAI-Scheduler/pull/779) [itsomri](https://github.com/itsomri)
+- Added clear error messages for jobs referencing missing or orphan queues, reporting via events and conditions [#820](https://github.com/NVIDIA/KAI-Scheduler/pull/820) [gshaibi](https://github.com/gshaibi)
+- Added rule selector for resource accounting prometheus [#818](https://github.com/NVIDIA/KAI-Scheduler/pull/818) [itsomri](https://github.com/itsomri)
+- Made accounting labels configurable [#818](https://github.com/NVIDIA/KAI-Scheduler/pull/818) [itsomri](https://github.com/itsomri)
+- Added support for Grove hierarchical topology constraints in PodGroup subgroups
+- Added support for n-level queue hierarchies [#858](https://github.com/NVIDIA/KAI-Scheduler/pull/858) [gshaibi](https://github.com/gshaibi)
+- Added labels and annotations propagation from topOwner in SkipTopOwner grouper [#861](https://github.com/NVIDIA/KAI-Scheduler/pull/861) [SiorMeir](https://github.com/siormeir)
+- Added scheduler name match conditions to admission webhooks to improve cluster stability
+- Add Gpu Dra claims and resource slices accounting for the purpose of resource management and quota guarantees. *** This change doesn't support shared gpu claims or gpu claims with FirstAvailable *** [#900](https://github.com/NVIDIA/KAI-Scheduler/pull/900) [davidLif](https://github.com/davidLif) 
+- Added DRA resources recording to snapshot [#830](https://github.com/NVIDIA/KAI-Scheduler/pull/830)
+- Temporarily Prevent device-plugin GPU pods on DRA-only nodes - until translation between device-plugin notation and DRA is implemented
+- Implemented subgroups for pytorchjobs [#935](https://github.com/NVIDIA/KAI-Scheduler/pull/935) [itsomri](https://github.com/itsomri)
+- Made KAI images distroless [#745](https://github.com/NVIDIA/KAI-Scheduler/pull/745) [dttung2905](https://github.com/dttung2905)
+- Allow setting empty gpuPodRuntimeClassName during helm install [#972](https://github.com/NVIDIA/KAI-Scheduler/pull/972) [steved](https://github.com/steved)
+- Created scale tests scenarios for running scale tests for KAI [#967](https://github.com/NVIDIA/KAI-Scheduler/pull/967)
+- Implemented block-level segmentation for pytorchjobs [#938](https://github.com/NVIDIA/KAI-Scheduler/pull/938) [itsomri](https://github.com/itsomri)
+- Added scale test environment setup script and updated service monitors for KAI scheduler [#1031](https://github.com/NVIDIA/KAI-Scheduler/pull/1031)
+- Implemented subgroups for leaderworkerset [#1046](https://github.com/NVIDIA/KAI-Scheduler/pull/1046) [davidLif](https://github.com/davidLif) 
+- Added discovery data to snapshot for more accurate debugging [#1047](https://github.com/NVIDIA/KAI-Scheduler/pull/1047) [itsomri](https://github.com/itsomri)
+- Implemented subgroup segmentation (with topology segment definitions) for leaderworkerset [#1058](https://github.com/NVIDIA/KAI-Scheduler/pull/10586) [davidLif](https://github.com/davidLif)
 
 ### Fixed
 - Fixed a bug where queue status did not reflect its podgroups resources correctly [#1049](https://github.com/NVIDIA/KAI-Scheduler/pull/1049)

pkg/scheduler/api/resource_info/resource_info.go

@@ -52,6 +52,9 @@ func NewResource(milliCPU float64, memory float64, gpus float64) *Resource {
 func ResourceFromResourceList(rList v1.ResourceList) *Resource {
 	r := EmptyResource()
 	for rName, rQuant := range rList {
+		if rQuant.IsZero() {
+			continue
+		}
 		switch rName {
 		case v1.ResourceCPU:
 			r.milliCpu += float64(rQuant.MilliValue())

pkg/scheduler/api/resource_info/resource_info_test.go

@@ -0,0 +1,42 @@
+// Copyright 2025 NVIDIA CORPORATION
+// SPDX-License-Identifier: Apache-2.0
+
+package resource_info
+
+import (
+	v1 "k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/api/resource"
+
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("Resource internal logic", func() {
+	Context("ResourceFromResourceList", func() {
+		It("should skip resources with zero quantity", func() {
+			resourceList := v1.ResourceList{
+				v1.ResourceCPU:    resource.MustParse("1"),
+				v1.ResourceMemory: resource.MustParse("5G"),
+				GPUResourceName:   resource.MustParse("1"),
+				v1.ResourceName("nvidia.com/mig-1g.24gb"): resource.MustParse("0"),
+				v1.ResourceName("nvidia.com/mig-2g.48gb"): resource.MustParse("0"),
+				v1.ResourceName("rdma/ib0"):               resource.MustParse("0"),
+			}
+
+			resource := ResourceFromResourceList(resourceList)
+
+			Expect(resource.milliCpu).To(Equal(float64(1000)))
+			Expect(resource.memory).To(Equal(float64(5000000000)))
+			Expect(resource.gpus).To(Equal(float64(1)))
+
+			scalarResources := resource.ScalarResources()
+			_, hasMig1g := scalarResources[v1.ResourceName("nvidia.com/mig-1g.24gb")]
+			_, hasMig2g := scalarResources[v1.ResourceName("nvidia.com/mig-2g.48gb")]
+			_, hasRdma := scalarResources[v1.ResourceName("rdma/ib0")]
+
+			Expect(hasMig1g).To(BeFalse())
+			Expect(hasMig2g).To(BeFalse())
+			Expect(hasRdma).To(BeFalse())
+		})
+	})
+})