kserve/pkg/apis/serving/v1beta1/inference_service.go

/*
Copyright 2021 The KServe Authors.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/

package v1beta1

import (
	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)

// InferenceServiceSpec is the top level type for this resource
type InferenceServiceSpec struct {
	// Predictor defines the model serving spec
	// +required
	Predictor PredictorSpec `json:"predictor"`
	// Explainer defines the model explanation service spec,
	// explainer service calls to predictor or transformer if it is specified.
	// +optional
	Explainer *ExplainerSpec `json:"explainer,omitempty"`
	// Transformer defines the pre/post processing before and after the predictor call,
	// transformer service calls to predictor service.
	// +optional
	Transformer *TransformerSpec `json:"transformer,omitempty"`
}

// StorageSpec defines a spec for an object in an object store
type StorageSpec struct {
	// The path to the object in the storage. Note that this path is relative to the storage URI.
	// +optional
	Path *string `json:"path,omitempty"`
	// Parameters to override the default storage credentials and config.
	// +optional
	Parameters *map[string]string `json:"parameters,omitempty"`
	// The Storage Key in the secret for this object.
	// +optional
	StorageKey *string `json:"key,omitempty"`
}

type LoggerStorageSpec struct {
	StorageSpec        `json:",inline"`
	ServiceAccountName *string `json:"serviceAccountName,omitempty"`
}

// LoggerType controls the scope of log publishing
// +kubebuilder:validation:Enum=all;request;response
type LoggerType string

// LoggerType Enum
const (
	// LogAll Logger mode to log both request and response
	LogAll LoggerType = "all"
	// LogRequest Logger mode to log only request
	LogRequest LoggerType = "request"
	// LogResponse Logger mode to log only response
	LogResponse LoggerType = "response"
)

// LoggerSpec specifies optional payload logging available for all components
type LoggerSpec struct {
	// URL to send logging events
	// +optional
	URL *string `json:"url,omitempty"`
	// Specifies the scope of the loggers. <br />
	// Valid values are: <br />
	// - "all" (default): log both request and response; <br />
	// - "request": log only request; <br />
	// - "response": log only response <br />
	// +optional
	Mode LoggerType `json:"mode,omitempty"`
	// Matched metadata HTTP headers for propagating to inference logger cloud events.
	// +optional
	MetadataHeaders []string `json:"metadataHeaders,omitempty"`
	// Matched inference service annotations for propagating to inference logger cloud events.
	// +optional
	MetadataAnnotations []string `json:"metadataAnnotations,omitempty"`
	// Specifies the storage location for the inference logger cloud events.
	// +optional
	Storage *LoggerStorageSpec `json:"storage,omitempty"`
	// URL of the log marshaller service that transforms log records before storage.
	// Defaults to the embedded JSON marshaller at http://localhost:9083/marshal.
	// +optional
	MarshallerURL *string `json:"marshallerUrl,omitempty"`
	// Number of log records per batch for blob storage. Defaults to 1 (immediate).
	// +optional
	BatchSize *int `json:"batchSize,omitempty"`
	// Max duration to wait before flushing a partial batch (e.g. "5s", "100ms").
	// Only used when BatchSize > 1. Defaults to "0" (no time-based flushing).
	// +optional
	BatchInterval *string `json:"batchInterval,omitempty"`
}

// MetricsBackend enum
// +kubebuilder:validation:Enum=prometheus;graphite
type MetricsBackend string

const (
	PrometheusBackend MetricsBackend = "prometheus"
	GraphiteBackend   MetricsBackend = "graphite"
)

// PodsMetricsBackend enum
// +kubebuilder:validation:Enum=opentelemetry
type PodsMetricsBackend string

const (
	OpenTelemetryBackend PodsMetricsBackend = "opentelemetry"
)

// Batcher specifies optional payload batching available for all components
type Batcher struct {
	// Specifies the max number of requests to trigger a batch
	// +optional
	MaxBatchSize *int `json:"maxBatchSize,omitempty"`
	// Specifies the max latency to trigger a batch
	// +optional
	MaxLatency *int `json:"maxLatency,omitempty"`
	// Specifies the timeout of a batch
	// +optional
	Timeout *int `json:"timeout,omitempty"`
}

// InferenceService is the Schema for the InferenceServices API
// +k8s:openapi-gen=true
// +genclient
// +kubebuilder:object:root=true
// +kubebuilder:subresource:status
// +kubebuilder:printcolumn:name="URL",type="string",JSONPath=".status.url"
// +kubebuilder:printcolumn:name="Ready",type="string",JSONPath=".status.conditions[?(@.type=='Ready')].status"
// +kubebuilder:printcolumn:name="Prev",type="integer",JSONPath=".status.components.predictor.traffic[?(@.tag=='prev')].percent"
// +kubebuilder:printcolumn:name="Latest",type="integer",JSONPath=".status.components.predictor.traffic[?(@.latestRevision==true)].percent"
// +kubebuilder:printcolumn:name="PrevRolledoutRevision",type="string",JSONPath=".status.components.predictor.traffic[?(@.tag=='prev')].revisionName"
// +kubebuilder:printcolumn:name="LatestReadyRevision",type="string",JSONPath=".status.components.predictor.traffic[?(@.latestRevision==true)].revisionName"
// +kubebuilder:printcolumn:name="Age",type="date",JSONPath=".metadata.creationTimestamp"
// +kubebuilder:resource:path=inferenceservices,shortName=isvc
// +kubebuilder:storageversion
type InferenceService struct {
	metav1.TypeMeta   `json:",inline"`
	metav1.ObjectMeta `json:"metadata,omitempty"`

	Spec InferenceServiceSpec `json:"spec,omitempty"`

	// +kubebuilder:pruning:PreserveUnknownFields
	Status InferenceServiceStatus `json:"status,omitempty"`
}

// InferenceServiceList contains a list of Service
// +k8s:openapi-gen=true
// +kubebuilder:object:root=true
type InferenceServiceList struct {
	metav1.TypeMeta `json:",inline"`
	metav1.ListMeta `json:"metadata,omitempty"`
	// +listType=set
	Items []InferenceService `json:"items"`
}

func init() {
	SchemeBuilder.Register(&InferenceService{}, &InferenceServiceList{})
}