Merge branch 'indent-namespaceselector' into helm-and-securitycontext

.golangci.yaml

@@ -11,7 +11,6 @@ linters-settings:
       - assignOp
       - captLocal
       - commentFormatting
-      - deprecatedComment
       - elseif
       - exitAfterDefer
       - ifElseChain
@@ -20,6 +19,24 @@ linters-settings:
   govet:
     enable:
       - nilness
+  gci:
+    # Section configuration to compare against.
+    # Section names are case-insensitive and may contain parameters in ().
+    # The default order of sections is `standard > default > custom > blank > dot > alias`,
+    # If `custom-order` is `true`, it follows the order of `sections` option.
+    # Default: ["standard", "default"]
+    sections:
+      - standard # Standard section: captures all standard packages.
+      - default # Default section: contains all imports that could not be matched to another section type.
+      - prefix(sigs.k8s.io/kueue) # Custom section: groups all imports with the specified Prefix.
+    # Skip generated files.
+    # Default: true
+    skip-generated: true
+    # Enable custom order of sections.
+    # If `true`, make the section order the same as the order of `sections`.
+    # Default: false
+    custom-order: true
+
 
 # Settings for enabling and disabling linters
 linters:
@@ -30,6 +47,7 @@ linters:
     - gocritic
     - goimports
     - govet
+    - loggercheck
     - misspell
     - unconvert
 
@@ -38,6 +56,12 @@ issues:
   # Which dirs to exclude: issues from them won't be reported
   exclude-dirs:
     - bin
+  # Excluding configuration per-path, per-linter, per-text and per-source
+  exclude-rules:
+    - linters:
+      - staticcheck
+      # TODO(#768): Drop when incrementing the API version.
+      text: "SA1019: constants.QueueAnnotation is deprecated"
   # Show all issues from a linter
   max-issues-per-linter: 0
   # Show all issues with the same text

Makefile

@@ -213,14 +213,15 @@ run-test-multikueue-e2e-%: FORCE
 	@echo Running multikueue e2e for k8s ${K8S_VERSION}
 	E2E_KIND_VERSION="kindest/node:v$(K8S_VERSION)" KIND_CLUSTER_NAME=$(KIND_CLUSTER_NAME) CREATE_KIND_CLUSTER=$(CREATE_KIND_CLUSTER) ARTIFACTS="$(ARTIFACTS)/$@" IMAGE_TAG=$(IMAGE_TAG) GINKGO_ARGS="$(GINKGO_ARGS)" JOBSET_VERSION=$(JOBSET_VERSION) ./hack/multikueue-e2e-test.sh
 
-SCALABILITY_RUNNER := $(ARTIFACTS)/scalability-runner
-.PHONY: scalability-runner
-scalability-runner:
-	$(GO_BUILD_ENV) $(GO_CMD) build -ldflags="$(LD_FLAGS)" -o $(SCALABILITY_RUNNER) test/scalability/runner/main.go
+SCALABILITY_RUNNER := $(PROJECT_DIR)/bin/performance-scheduler-runner
+.PHONY: performance-scheduler-runner
+performance-scheduler-runner:
+	$(GO_BUILD_ENV) $(GO_CMD) build -ldflags="$(LD_FLAGS)" -o $(SCALABILITY_RUNNER) test/performance/scheduler/runner/main.go
 
+MINIMALKUEUE_RUNNER := $(PROJECT_DIR)/bin/minimalkueue
 .PHONY: minimalkueue
 minimalkueue: 
-	$(GO_BUILD_ENV) $(GO_CMD) build -ldflags="$(LD_FLAGS)" -o  $(ARTIFACTS)/minimalkueue test/scalability/minimalkueue/main.go
+	$(GO_BUILD_ENV) $(GO_CMD) build -ldflags="$(LD_FLAGS)" -o $(MINIMALKUEUE_RUNNER) test/performance/scheduler/minimalkueue/main.go
 
 ifdef SCALABILITY_CPU_PROFILE
 SCALABILITY_EXTRA_ARGS += --withCPUProfile=true
@@ -230,38 +231,51 @@ ifdef SCALABILITY_KUEUE_LOGS
 SCALABILITY_EXTRA_ARGS +=  --withLogs=true --logToFile=true
 endif
 
-SCALABILITY_GENERATOR_CONFIG ?= $(PROJECT_DIR)/test/scalability/default_generator_config.yaml
+ifdef SCALABILITY_SCRAPE_INTERVAL
+SCALABILITY_SCRAPE_ARGS +=  --metricsScrapeInterval=$(SCALABILITY_SCRAPE_INTERVAL)
+endif
+
+ifdef SCALABILITY_SCRAPE_URL
+SCALABILITY_SCRAPE_ARGS +=  --metricsScrapeURL=$(SCALABILITY_SCRAPE_URL)
+endif
 
-SCALABILITY_RUN_DIR := $(ARTIFACTS)/run-scalability
-.PHONY: run-scalability
-run-scalability: envtest scalability-runner minimalkueue
+SCALABILITY_GENERATOR_CONFIG ?= $(PROJECT_DIR)/test/performance/scheduler/default_generator_config.yaml
+
+SCALABILITY_RUN_DIR := $(ARTIFACTS)/run-performance-scheduler
+.PHONY: run-performance-scheduler
+run-performance-scheduler: envtest performance-scheduler-runner minimalkueue
 	mkdir -p $(SCALABILITY_RUN_DIR)
 	KUBEBUILDER_ASSETS="$(shell $(ENVTEST) use $(ENVTEST_K8S_VERSION) -p path)" \
 	$(SCALABILITY_RUNNER) \
 		--o $(SCALABILITY_RUN_DIR) \
 		--crds=$(PROJECT_DIR)/config/components/crd/bases \
 		--generatorConfig=$(SCALABILITY_GENERATOR_CONFIG) \
-		--minimalKueue=$(ARTIFACTS)/minimalkueue $(SCALABILITY_EXTRA_ARGS)
+		--minimalKueue=$(MINIMALKUEUE_RUNNER) $(SCALABILITY_EXTRA_ARGS) $(SCALABILITY_SCRAPE_ARGS)
 
-.PHONY: test-scalability
-test-scalability: gotestsum run-scalability
-	$(GOTESTSUM) --junitfile $(ARTIFACTS)/junit.xml -- $(GO_TEST_FLAGS) ./test/scalability/checker  \
+.PHONY: test-performance-scheduler
+test-performance-scheduler: gotestsum run-performance-scheduler
+	$(GOTESTSUM) --junitfile $(ARTIFACTS)/junit.xml -- $(GO_TEST_FLAGS) ./test/performance/scheduler/checker  \
+		--summary=$(SCALABILITY_RUN_DIR)/summary.yaml \
 		--cmdStats=$(SCALABILITY_RUN_DIR)/minimalkueue.stats.yaml \
-		--range=$(PROJECT_DIR)/test/scalability/default_rangespec.yaml
+		--range=$(PROJECT_DIR)/test/performance/scheduler/default_rangespec.yaml
 
-.PHONY: run-scalability-in-cluster
-run-scalability-in-cluster: envtest scalability-runner
-	mkdir -p $(ARTIFACTS)/run-scalability-in-cluster
+.PHONY: run-performance-scheduler-in-cluster
+run-performance-scheduler-in-cluster: envtest performance-scheduler-runner
+	mkdir -p $(ARTIFACTS)/run-performance-scheduler-in-cluster
 	KUBEBUILDER_ASSETS="$(shell $(ENVTEST) use $(ENVTEST_K8S_VERSION) -p path)" \
 	$(SCALABILITY_RUNNER) \
-		--o $(ARTIFACTS)/run-scalability-in-cluster \
+		--o $(ARTIFACTS)/run-performance-scheduler-in-cluster \
 		--generatorConfig=$(SCALABILITY_GENERATOR_CONFIG) \
-		--qps=1000 --burst=2000 --timeout=15m
+		--qps=1000 --burst=2000 --timeout=15m $(SCALABILITY_SCRAPE_ARGS)
 
 .PHONY: ci-lint
 ci-lint: golangci-lint
 	$(GOLANGCI_LINT) run --timeout 15m0s
 
+.PHONY: lint-fix
+lint-fix: golangci-lint
+	$(GOLANGCI_LINT) run --fix --timeout 15m0s
+
 .PHONY: verify
 verify: gomod-verify ci-lint fmt-verify shell-lint toc-verify manifests generate update-helm generate-apiref prepare-release-branch
 	git --no-pager diff --exit-code config/components apis charts/kueue/templates client-go site/
@@ -497,4 +511,4 @@ cluster-autoscaler-crd:
 
 .PHONY: generate-apiref
 generate-apiref: genref
-	cd  $(PROJECT_DIR)/site/genref/ && $(GENREF)  -o $(PROJECT_DIR)/site/content/en/docs/reference
+	cd $(PROJECT_DIR)/hack/genref/ && $(GENREF) -o $(PROJECT_DIR)/site/content/en/docs/reference

apis/config/v1beta1/configuration_types.go

@@ -238,16 +238,24 @@ type RequeuingStrategy struct {
 	// Once the number is reached, the workload is deactivated (`.spec.activate`=`false`).
 	// When it is null, the workloads will repeatedly and endless re-queueing.
 	//
-	// Every backoff duration is about "1.41284738^(n-1)+Rand" where the "n" represents the "workloadStatus.requeueState.count",
-	// and the "Rand" represents the random jitter. During this time, the workload is taken as an inadmissible and
+	// Every backoff duration is about "b*2^(n-1)+Rand" where:
+	// - "b" represents the base set by "BackoffBaseSeconds" parameter,
+	// - "n" represents the "workloadStatus.requeueState.count",
+	// - "Rand" represents the random jitter.
+	// During this time, the workload is taken as an inadmissible and
 	// other workloads will have a chance to be admitted.
-	// For example, when the "waitForPodsReady.timeout" is the default, the workload deactivation time is as follows:
-	//   {backoffLimitCount, workloadDeactivationSeconds}
-	//     ~= {1, 601}, {2, 902}, ...,{5, 1811}, ...,{10, 3374}, ...,{20, 8730}, ...,{30, 86400(=24 hours)}, ...
+	// By default, the consecutive requeue delays are around: (10s, 20s, 40s, ...).
 	//
 	// Defaults to null.
 	// +optional
 	BackoffLimitCount *int32 `json:"backoffLimitCount,omitempty"`
+
+	// BackoffBaseSeconds defines the base for the exponential backoff for
+	// re-queuing an evicted workload.
+	//
+	// Defaults to 10.
+	// +optional
+	BackoffBaseSeconds *int32 `json:"backoffBaseSeconds,omitempty"`
 }
 
 type RequeuingTimestamp string
@@ -299,6 +307,9 @@ type Integrations struct {
 	//  - "kubeflow.org/xgboostjob"
 	//  - "pod"
 	Frameworks []string `json:"frameworks,omitempty"`
+	// List of GroupVersionKinds that are managed for Kueue by external controllers;
+	// the expected format is `Kind.version.group.com`.
+	ExternalFrameworks []string `json:"externalFrameworks,omitempty"`
 	// PodOptions defines kueue controller behaviour for pod objects
 	PodOptions *PodIntegrationOptions `json:"podOptions,omitempty"`
 

apis/config/v1beta1/defaults.go

@@ -47,6 +47,7 @@ const (
 	DefaultMultiKueueGCInterval                         = time.Minute
 	DefaultMultiKueueOrigin                             = "multikueue"
 	DefaultMultiKueueWorkerLostTimeout                  = 15 * time.Minute
+	DefaultRequeuingBackoffBaseSeconds                  = 10
 )
 
 func getOperatorNamespace() string {
@@ -121,10 +122,14 @@ func SetDefaults_Configuration(cfg *Configuration) {
 			}
 			cfg.WaitForPodsReady.BlockAdmission = &defaultBlockAdmission
 		}
-		if cfg.WaitForPodsReady.RequeuingStrategy == nil || cfg.WaitForPodsReady.RequeuingStrategy.Timestamp == nil {
-			cfg.WaitForPodsReady.RequeuingStrategy = &RequeuingStrategy{
-				Timestamp: ptr.To(EvictionTimestamp),
-			}
+		if cfg.WaitForPodsReady.RequeuingStrategy == nil {
+			cfg.WaitForPodsReady.RequeuingStrategy = &RequeuingStrategy{}
+		}
+		if cfg.WaitForPodsReady.RequeuingStrategy.Timestamp == nil {
+			cfg.WaitForPodsReady.RequeuingStrategy.Timestamp = ptr.To(EvictionTimestamp)
+		}
+		if cfg.WaitForPodsReady.RequeuingStrategy.BackoffBaseSeconds == nil {
+			cfg.WaitForPodsReady.RequeuingStrategy.BackoffBaseSeconds = ptr.To[int32](DefaultRequeuingBackoffBaseSeconds)
 		}
 	}
 	if cfg.Integrations == nil {

apis/config/v1beta1/defaults_test.go

@@ -364,7 +364,8 @@ func TestSetDefaults_Configuration(t *testing.T) {
 					BlockAdmission: ptr.To(true),
 					Timeout:        &podsReadyTimeoutTimeout,
 					RequeuingStrategy: &RequeuingStrategy{
-						Timestamp: ptr.To(EvictionTimestamp),
+						Timestamp:          ptr.To(EvictionTimestamp),
+						BackoffBaseSeconds: ptr.To[int32](DefaultRequeuingBackoffBaseSeconds),
 					},
 				},
 				Namespace:         ptr.To(DefaultNamespace),
@@ -393,7 +394,8 @@ func TestSetDefaults_Configuration(t *testing.T) {
 					BlockAdmission: ptr.To(false),
 					Timeout:        &podsReadyTimeoutTimeout,
 					RequeuingStrategy: &RequeuingStrategy{
-						Timestamp: ptr.To(EvictionTimestamp),
+						Timestamp:          ptr.To(EvictionTimestamp),
+						BackoffBaseSeconds: ptr.To[int32](DefaultRequeuingBackoffBaseSeconds),
 					},
 				},
 				Namespace:         ptr.To(DefaultNamespace),
@@ -413,7 +415,8 @@ func TestSetDefaults_Configuration(t *testing.T) {
 					Enable:  true,
 					Timeout: &podsReadyTimeoutOverwrite,
 					RequeuingStrategy: &RequeuingStrategy{
-						Timestamp: ptr.To(CreationTimestamp),
+						Timestamp:          ptr.To(CreationTimestamp),
+						BackoffBaseSeconds: ptr.To[int32](63),
 					},
 				},
 				InternalCertManagement: &InternalCertManagement{
@@ -426,7 +429,8 @@ func TestSetDefaults_Configuration(t *testing.T) {
 					BlockAdmission: ptr.To(true),
 					Timeout:        &podsReadyTimeoutOverwrite,
 					RequeuingStrategy: &RequeuingStrategy{
-						Timestamp: ptr.To(CreationTimestamp),
+						Timestamp:          ptr.To(CreationTimestamp),
+						BackoffBaseSeconds: ptr.To[int32](63),
 					},
 				},
 				Namespace:         ptr.To(DefaultNamespace),

apis/kueue/v1beta1/admissioncheck_types.go

@@ -104,6 +104,9 @@ const (
 	// Having multiple AdmissionChecks managed by the same controller where at least one has this condition
 	// set to true will cause the ClusterQueue to be marked as Inactive.
 	AdmissionChecksSingleInstanceInClusterQueue string = "SingleInstanceInClusterQueue"
+
+	// FlavorIndependentAdmissionCheck indicates if the AdmissionCheck cannot be applied at ResourceFlavor level.
+	FlavorIndependentAdmissionCheck string = "FlavorIndependent"
 )
 
 // +genclient