[Core] Improve model adapter controller by handling existing resources (

#54) * Update ModelAdapter API definition 1. Clean up the conditions and only leave phase and []metav1.Condition 2. Enrich the phase and condition type * Improve model adapter controller implementation 1. Update phase and conditions in the workflow 2. Initially support finalizer 3. Handle resource existing cases (status.instance, service & endpointslice etc)
vllm-project · Jul 29, 2024 · adeb124 · adeb124
1 parent a5743da
commit adeb124
Show file tree

Hide file tree

Showing 4 changed files with 384 additions and 182 deletions.
diff --git a/api/model/v1alpha1/modeladapter_types.go b/api/model/v1alpha1/modeladapter_types.go
@@ -17,7 +17,6 @@ limitations under the License.
 package v1alpha1
 
 import (
-	corev1 "k8s.io/api/core/v1"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 )
 
@@ -49,61 +48,42 @@ const (
 	ModelAdapterPending ModelAdapterPhase = "Pending"
 	// ModelAdapterScheduling means the ModelAdapter is pending scheduling
 	ModelAdapterScheduling ModelAdapterPhase = "Scheduling"
-	// ModelAdapterBinding means the controller starts to load ModelAdapter on a selected pod
+	// ModelAdapterBinding means the controller loads ModelAdapter on a selected pod
 	ModelAdapterBinding ModelAdapterPhase = "Binding"
-	// ModelAdapterConfiguring means the controller starts to configure the service and endpoint for ModelAdapter
-	ModelAdapterConfiguring ModelAdapterPhase = "Configuring"
 	// ModelAdapterRunning means ModelAdapter has been running on the pod
 	ModelAdapterRunning ModelAdapterPhase = "Running"
 	// ModelAdapterFailed means ModelAdapter has terminated in a failure
 	ModelAdapterFailed ModelAdapterPhase = "Failed"
-	// ModelAdapterScaling means ModelAdapter is scaling, could be scaling in or out
+	// ModelAdapterScaling means ModelAdapter is scaling, could be scaling in or out. won't be enabled until we allow multiple replicas
 	ModelAdapterScaling ModelAdapterPhase = "Scaling"
 )
 
 // ModelAdapterStatus defines the observed state of ModelAdapter
 type ModelAdapterStatus struct {
-
 	// Phase is a simple, high-level summary of where the ModelAdapter is in its lifecycle
+	// Phase maps to latest status.conditions.type
 	// +optional
 	Phase ModelAdapterPhase `json:"phase,omitempty"`
-	// Conditions represents the latest available observations of an object's state
+	// Conditions represents the observation of a model adapter's current state.
 	// +patchMergeKey=type
 	// +patchStrategy=merge
 	// +optional
-	Conditions []ModelAdapterCondition `json:"conditions,omitempty"`
-	// LastTransitionTime is the time the last Phase transitioned to the current one
-	// +optional
-	LastTransitionTime *metav1.Time `json:"lastTransitionTime,omitempty"`
-	// Reason is a unique, one-word, CamelCase reason for the phase's last transition
-	// +optional
-	Reason string `json:"reason,omitempty"`
-	// Message is a human-readable message indicating details about the last transition
-	// +optional
-	Message string `json:"message,omitempty"`
+	Conditions []metav1.Condition `json:"conditions,omitempty"`
 	// Instances lists all pod instances of ModelAdapter
 	// +optional
 	Instances []string `json:"instances,omitempty"`
 }
 
 type ModelAdapterConditionType string
 
-// ModelAdapterCondition contains details for the current condition of this ModelAdapter
-type ModelAdapterCondition struct {
-	// Type is the type of the condition
-	Type ModelAdapterConditionType `json:"type"`
-	// Status is the status of the condition
-	Status corev1.ConditionStatus `json:"status"`
-	// LastTransitionTime is the time the condition last transitioned from one status to another
-	// +optional
-	LastTransitionTime metav1.Time `json:"lastTransitionTime,omitempty"`
-	// Reason is a unique, one-word, CamelCase reason for the condition's last transition
-	// +optional
-	Reason string `json:"reason,omitempty"`
-	// Message is a human-readable message indicating details about the transition
-	// +optional
-	Message string `json:"message,omitempty"`
-}
+const (
+	ModelAdapterConditionTypeInitialized     ModelAdapterConditionType = "Initialized"
+	ModelAdapterConditionTypeSelectorMatched ModelAdapterConditionType = "SelectorMatched"
+	ModelAdapterConditionTypeScheduled       ModelAdapterConditionType = "Scheduled"
+	ModelAdapterConditionTypeResourceCreated ModelAdapterConditionType = "ResourceCreated"
+	ModelAdapterConditionReady               ModelAdapterConditionType = "Ready"
+	ModelAdapterConditionCleanup             ModelAdapterConditionType = "Cleanup"
+)
 
 //+kubebuilder:object:root=true
 //+kubebuilder:subresource:status

diff --git a/api/model/v1alpha1/zz_generated.deepcopy.go b/api/model/v1alpha1/zz_generated.deepcopy.go
diff --git a/config/crd/bases/model.aibrix.ai_modeladapters.yaml b/config/crd/bases/model.aibrix.ai_modeladapters.yaml
@@ -103,32 +103,72 @@ spec:
             description: ModelAdapterStatus defines the observed state of ModelAdapter
             properties:
               conditions:
-                description: Conditions represents the latest available observations
-                  of an object's state
+                description: Conditions represents the observation of a model adapter's
+                  current state.
                 items:
-                  description: ModelAdapterCondition contains details for the current
-                    condition of this ModelAdapter
+                  description: "Condition contains details for one aspect of the current
+                    state of this API Resource.\n---\nThis struct is intended for
+                    direct use as an array at the field path .status.conditions.  For
+                    example,\n\n\n\ttype FooStatus struct{\n\t    // Represents the
+                    observations of a foo's current state.\n\t    // Known .status.conditions.type
+                    are: \"Available\", \"Progressing\", and \"Degraded\"\n\t    //
+                    +patchMergeKey=type\n\t    // +patchStrategy=merge\n\t    // +listType=map\n\t
+                    \   // +listMapKey=type\n\t    Conditions []metav1.Condition `json:\"conditions,omitempty\"
+                    patchStrategy:\"merge\" patchMergeKey:\"type\" protobuf:\"bytes,1,rep,name=conditions\"`\n\n\n\t
+                    \   // other fields\n\t}"
                   properties:
                     lastTransitionTime:
-                      description: LastTransitionTime is the time the condition last
-                        transitioned from one status to another
+                      description: |-
+                        lastTransitionTime is the last time the condition transitioned from one status to another.
+                        This should be when the underlying condition changed.  If that is not known, then using the time when the API field changed is acceptable.
                       format: date-time
                       type: string
                     message:
-                      description: Message is a human-readable message indicating
-                        details about the transition
+                      description: |-
+                        message is a human readable message indicating details about the transition.
+                        This may be an empty string.
+                      maxLength: 32768
                       type: string
+                    observedGeneration:
+                      description: |-
+                        observedGeneration represents the .metadata.generation that the condition was set based upon.
+                        For instance, if .metadata.generation is currently 12, but the .status.conditions[x].observedGeneration is 9, the condition is out of date
+                        with respect to the current state of the instance.
+                      format: int64
+                      minimum: 0
+                      type: integer
                     reason:
-                      description: Reason is a unique, one-word, CamelCase reason
-                        for the condition's last transition
+                      description: |-
+                        reason contains a programmatic identifier indicating the reason for the condition's last transition.
+                        Producers of specific condition types may define expected values and meanings for this field,
+                        and whether the values are considered a guaranteed API.
+                        The value should be a CamelCase string.
+                        This field may not be empty.
+                      maxLength: 1024
+                      minLength: 1
+                      pattern: ^[A-Za-z]([A-Za-z0-9_,:]*[A-Za-z0-9_])?$
                       type: string
                     status:
-                      description: Status is the status of the condition
+                      description: status of the condition, one of True, False, Unknown.
+                      enum:
+                      - "True"
+                      - "False"
+                      - Unknown
                       type: string
                     type:
-                      description: Type is the type of the condition
+                      description: |-
+                        type of condition in CamelCase or in foo.example.com/CamelCase.
+                        ---
+                        Many .condition.type values are consistent across resources like Available, but because arbitrary conditions can be
+                        useful (see .node.status.conditions), the ability to deconflict is important.
+                        The regex it matches is (dns1123SubdomainFmt/)?(qualifiedNameFmt)
+                      maxLength: 316
+                      pattern: ^([a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*/)?(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9])$
                       type: string
                   required:
+                  - lastTransitionTime
+                  - message
+                  - reason
                   - status
                   - type
                   type: object
@@ -138,22 +178,10 @@ spec:
                 items:
                   type: string
                 type: array
-              lastTransitionTime:
-                description: LastTransitionTime is the time the last Phase transitioned
-                  to the current one
-                format: date-time
-                type: string
-              message:
-                description: Message is a human-readable message indicating details
-                  about the last transition
-                type: string
               phase:
-                description: Phase is a simple, high-level summary of where the ModelAdapter
-                  is in its lifecycle
-                type: string
-              reason:
-                description: Reason is a unique, one-word, CamelCase reason for the
-                  phase's last transition
+                description: |-
+                  Phase is a simple, high-level summary of where the ModelAdapter is in its lifecycle
+                  Phase maps to latest status.conditions.type
                 type: string
             type: object
         type: object