feat(k8s): add Pool rolling update maxUnavailable, close #610

kubernetes/Makefile

@@ -124,6 +124,7 @@ test: manifests generate fmt vet setup-envtest ## Run tests.
 KIND_CLUSTER ?= sandbox-k8s-test-e2e
 KIND_K8S_VERSION ?= v1.22.4
 GINKGO_ARGS ?=
+E2E_TIMEOUT ?= 30m
 
 .PHONY: install-kind
 install-kind: ## Install Kind using go install if not already installed
@@ -148,15 +149,15 @@ setup-test-e2e: install-kind ## Set up a Kind cluster for e2e tests if it does n
 .PHONY: test-e2e-main
 test-e2e-main: setup-test-e2e manifests generate fmt vet 
 	CONTROLLER_IMG=$(CONTROLLER_IMG) TASK_EXECUTOR_IMG=$(TASK_EXECUTOR_IMG) \
-		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e/ -v -ginkgo.v $(GINKGO_ARGS)
+		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e/ -v -ginkgo.v -timeout $(E2E_TIMEOUT) $(GINKGO_ARGS)
 	$(MAKE) cleanup-test-e2e
 
 .PHONY: test-e2e
 test-e2e: setup-test-e2e manifests generate fmt vet ## Run the e2e tests. Expected an isolated environment using Kind. Use GINKGO_ARGS to pass additional arguments.
 	CONTROLLER_IMG=$(CONTROLLER_IMG) TASK_EXECUTOR_IMG=$(TASK_EXECUTOR_IMG) \
-		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e/ -v -ginkgo.v $(GINKGO_ARGS)
+		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e/ -v -ginkgo.v -timeout $(E2E_TIMEOUT) $(GINKGO_ARGS)
 	CONTROLLER_IMG=$(CONTROLLER_IMG) TASK_EXECUTOR_IMG=$(TASK_EXECUTOR_IMG) \
-		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e_task/ -v -ginkgo.v $(GINKGO_ARGS)
+		KIND_CLUSTER=$(KIND_CLUSTER) go test ./test/e2e_task/ -v -ginkgo.v -timeout $(E2E_TIMEOUT) $(GINKGO_ARGS)
 	$(MAKE) cleanup-test-e2e
 	$(MAKE) test-gvisor CONTROLLER_IMG=$(CONTROLLER_IMG) TASK_EXECUTOR_IMG=$(TASK_EXECUTOR_IMG)
 	$(MAKE) cleanup-gvisor

kubernetes/apis/sandbox/v1alpha1/pool_types.go

@@ -36,6 +36,9 @@ type PoolSpec struct {
 	// ScaleStrategy controls the scaling behavior.
 	// +optional
 	ScaleStrategy *ScaleStrategy `json:"scaleStrategy,omitempty"`
+	// UpdateStrategy controls how pool pods are updated when the template changes.
+	// +optional
+	UpdateStrategy *UpdateStrategy `json:"updateStrategy,omitempty"`
 }
 
 type CapacitySpec struct {
@@ -66,6 +69,15 @@ type ScaleStrategy struct {
 	MaxUnavailable *intstr.IntOrString `json:"maxUnavailable,omitempty"`
 }
 
+// UpdateStrategy controls how pool pods are updated when the pool template changes.
+type UpdateStrategy struct {
+	// MaxUnavailable is the maximum number of pods that can be unavailable during an update.
+	// Can be an absolute number (ex: 5) or a percentage of desired pods (ex: "20%").
+	// Defaults to 25%.
+	// +optional
+	MaxUnavailable *intstr.IntOrString `json:"maxUnavailable,omitempty"`
+}
+
 // PoolStatus defines the observed state of Pool.
 type PoolStatus struct {
 	// ObservedGeneration is the most recent generation observed for this BatchSandbox. It corresponds to the
@@ -79,6 +91,8 @@ type PoolStatus struct {
 	Allocated int32 `json:"allocated"`
 	// Available is the number of nodes currently available in the pool.
 	Available int32 `json:"available"`
+	// Updated is the number of nodes that have been updated to the latest revision.
+	Updated int32 `json:"updated,omitempty"`
 }
 
 // +genclient
@@ -88,6 +102,8 @@ type PoolStatus struct {
 // +kubebuilder:printcolumn:name="TOTAL",type="integer",JSONPath=".status.total",description="The number of all nodes in pool."
 // +kubebuilder:printcolumn:name="ALLOCATED",type="integer",JSONPath=".status.allocated",description="The number of allocated nodes in pool."
 // +kubebuilder:printcolumn:name="AVAILABLE",type="integer",JSONPath=".status.available",description="The number of available nodes in pool."
+// +kubebuilder:printcolumn:name="UPDATED",type="integer",JSONPath=".status.updated",description="The number of nodes updated to the latest revision."
+// +kubebuilder:printcolumn:name="AGE",type="date",JSONPath=".metadata.creationTimestamp"
 // Pool is the Schema for the pools API.
 type Pool struct {
 	metav1.TypeMeta   `json:",inline"`

kubernetes/config/crd/bases/sandbox.opensandbox.io_pools.yaml

@@ -27,6 +27,13 @@ spec:
       jsonPath: .status.available
       name: AVAILABLE
       type: integer
+    - description: The number of nodes updated to the latest revision.
+      jsonPath: .status.updated
+      name: UPDATED
+      type: integer
+    - jsonPath: .metadata.creationTimestamp
+      name: AGE
+      type: date
     name: v1alpha1
     schema:
       openAPIV3Schema:
@@ -100,6 +107,20 @@ spec:
               template:
                 description: Pod Template used to create pre-warmed nodes in the pool.
                 x-kubernetes-preserve-unknown-fields: true
+              updateStrategy:
+                description: UpdateStrategy controls how pool pods are updated when
+                  the template changes.
+                properties:
+                  maxUnavailable:
+                    anyOf:
+                    - type: integer
+                    - type: string
+                    description: |-
+                      MaxUnavailable is the maximum number of pods that can be unavailable during an update.
+                      Can be an absolute number (ex: 5) or a percentage of desired pods (ex: "20%").
+                      Defaults to 25%.
+                    x-kubernetes-int-or-string: true
+                type: object
             required:
             - capacitySpec
             type: object
@@ -129,6 +150,11 @@ spec:
                 description: Total is the total number of nodes in the pool.
                 format: int32
                 type: integer
+              updated:
+                description: Updated is the number of nodes that have been updated
+                  to the latest revision.
+                format: int32
+                type: integer
             required:
             - allocated
             - available