feat(routing): 引入多执行池智能调度
将 Worker 发现、路由画像、容量与执行传输抽象为平台无关接口，新增 Kubernetes 和静态容量适配器，并以 shadow 模式接入生产配置。 实现网络与容量评分、路由防抖、池队列、同步 Worker 租约、一次性执行令牌，以及提交状态不明时禁止重复分配的安全语义。 新增 0105 兼容迁移、管理接口、指标、OpenAPI 和回归测试。已执行全量 Go 测试、go vet、OpenAPI、迁移安全、Compose 与 Kustomize 验证。
This commit is contained in:
@@ -2082,6 +2082,36 @@
|
||||
}
|
||||
}
|
||||
},
|
||||
"/api/admin/runtime/execution-pools": {
|
||||
"get": {
|
||||
"security": [
|
||||
{
|
||||
"BearerAuth": []
|
||||
}
|
||||
],
|
||||
"produces": [
|
||||
"application/json"
|
||||
],
|
||||
"tags": [
|
||||
"Admin Runtime"
|
||||
],
|
||||
"summary": "查询逻辑执行池、Worker 容量和上游路由健康状态",
|
||||
"responses": {
|
||||
"200": {
|
||||
"description": "OK",
|
||||
"schema": {
|
||||
"$ref": "#/definitions/httpapi.AdminExecutionPoolResponse"
|
||||
}
|
||||
},
|
||||
"500": {
|
||||
"description": "Internal Server Error",
|
||||
"schema": {
|
||||
"$ref": "#/definitions/httpapi.ErrorEnvelope"
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
"/api/admin/runtime/model-rate-limits": {
|
||||
"get": {
|
||||
"security": [
|
||||
@@ -10699,6 +10729,177 @@
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminExecutionPool": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"capabilities": {
|
||||
"type": "object",
|
||||
"additionalProperties": {}
|
||||
},
|
||||
"capacity": {
|
||||
"$ref": "#/definitions/httpapi.AdminExecutionPoolCapacity"
|
||||
},
|
||||
"labels": {
|
||||
"type": "object",
|
||||
"additionalProperties": {
|
||||
"type": "string"
|
||||
}
|
||||
},
|
||||
"poolId": {
|
||||
"type": "string"
|
||||
},
|
||||
"state": {
|
||||
"type": "string"
|
||||
},
|
||||
"workers": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"$ref": "#/definitions/httpapi.AdminExecutionPoolWorker"
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminExecutionPoolCapacity": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"activeTasks": {
|
||||
"type": "integer"
|
||||
},
|
||||
"allocated": {
|
||||
"type": "integer"
|
||||
},
|
||||
"critical": {
|
||||
"type": "boolean"
|
||||
},
|
||||
"estimatedQueueWaitMs": {
|
||||
"type": "integer"
|
||||
},
|
||||
"heavyCapacity": {
|
||||
"type": "integer"
|
||||
},
|
||||
"heavyTasks": {
|
||||
"type": "integer"
|
||||
},
|
||||
"queueDepth": {
|
||||
"type": "integer"
|
||||
},
|
||||
"resourceHeadroom": {
|
||||
"type": "number"
|
||||
},
|
||||
"safeCapacity": {
|
||||
"type": "integer"
|
||||
},
|
||||
"sampledAt": {
|
||||
"type": "string"
|
||||
},
|
||||
"stability": {
|
||||
"type": "number"
|
||||
},
|
||||
"workerCount": {
|
||||
"type": "integer"
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminExecutionPoolResponse": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"capturedAt": {
|
||||
"type": "string"
|
||||
},
|
||||
"pools": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"$ref": "#/definitions/httpapi.AdminExecutionPool"
|
||||
}
|
||||
},
|
||||
"routeHealth": {
|
||||
"type": "array",
|
||||
"items": {
|
||||
"$ref": "#/definitions/httpapi.AdminRouteHealth"
|
||||
}
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminExecutionPoolWorker": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"activeTasks": {
|
||||
"type": "integer"
|
||||
},
|
||||
"allocated": {
|
||||
"type": "integer"
|
||||
},
|
||||
"capabilities": {
|
||||
"type": "object",
|
||||
"additionalProperties": {}
|
||||
},
|
||||
"heartbeatAt": {
|
||||
"type": "string"
|
||||
},
|
||||
"instanceId": {
|
||||
"type": "string"
|
||||
},
|
||||
"pressureState": {
|
||||
"type": "string"
|
||||
},
|
||||
"protocolVersion": {
|
||||
"type": "string"
|
||||
},
|
||||
"revision": {
|
||||
"type": "string"
|
||||
},
|
||||
"safeCapacity": {
|
||||
"type": "integer"
|
||||
},
|
||||
"workerId": {
|
||||
"type": "string"
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminRouteHealth": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"connectTlsP95Ms": {
|
||||
"type": "integer"
|
||||
},
|
||||
"consecutiveFailures": {
|
||||
"type": "integer"
|
||||
},
|
||||
"consecutiveSuccesses": {
|
||||
"type": "integer"
|
||||
},
|
||||
"expiresAt": {
|
||||
"type": "string"
|
||||
},
|
||||
"firstByteP95Ms": {
|
||||
"type": "integer"
|
||||
},
|
||||
"jitterP95Ms": {
|
||||
"type": "integer"
|
||||
},
|
||||
"poolId": {
|
||||
"type": "string"
|
||||
},
|
||||
"routeProfileKey": {
|
||||
"type": "string"
|
||||
},
|
||||
"sampleCount": {
|
||||
"type": "integer"
|
||||
},
|
||||
"sampledAt": {
|
||||
"type": "string"
|
||||
},
|
||||
"state": {
|
||||
"type": "string"
|
||||
},
|
||||
"successRate": {
|
||||
"type": "number"
|
||||
},
|
||||
"uploadBytesPerSecond": {
|
||||
"type": "number"
|
||||
}
|
||||
}
|
||||
},
|
||||
"httpapi.AdminTaskListResponse": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
@@ -17072,6 +17273,9 @@
|
||||
"podUid": {
|
||||
"type": "string"
|
||||
},
|
||||
"poolId": {
|
||||
"type": "string"
|
||||
},
|
||||
"preparingTasks": {
|
||||
"type": "integer"
|
||||
},
|
||||
|
||||
@@ -72,6 +72,119 @@ definitions:
|
||||
$ref: '#/definitions/store.AccessRule'
|
||||
type: array
|
||||
type: object
|
||||
httpapi.AdminExecutionPool:
|
||||
properties:
|
||||
capabilities:
|
||||
additionalProperties: {}
|
||||
type: object
|
||||
capacity:
|
||||
$ref: '#/definitions/httpapi.AdminExecutionPoolCapacity'
|
||||
labels:
|
||||
additionalProperties:
|
||||
type: string
|
||||
type: object
|
||||
poolId:
|
||||
type: string
|
||||
state:
|
||||
type: string
|
||||
workers:
|
||||
items:
|
||||
$ref: '#/definitions/httpapi.AdminExecutionPoolWorker'
|
||||
type: array
|
||||
type: object
|
||||
httpapi.AdminExecutionPoolCapacity:
|
||||
properties:
|
||||
activeTasks:
|
||||
type: integer
|
||||
allocated:
|
||||
type: integer
|
||||
critical:
|
||||
type: boolean
|
||||
estimatedQueueWaitMs:
|
||||
type: integer
|
||||
heavyCapacity:
|
||||
type: integer
|
||||
heavyTasks:
|
||||
type: integer
|
||||
queueDepth:
|
||||
type: integer
|
||||
resourceHeadroom:
|
||||
type: number
|
||||
safeCapacity:
|
||||
type: integer
|
||||
sampledAt:
|
||||
type: string
|
||||
stability:
|
||||
type: number
|
||||
workerCount:
|
||||
type: integer
|
||||
type: object
|
||||
httpapi.AdminExecutionPoolResponse:
|
||||
properties:
|
||||
capturedAt:
|
||||
type: string
|
||||
pools:
|
||||
items:
|
||||
$ref: '#/definitions/httpapi.AdminExecutionPool'
|
||||
type: array
|
||||
routeHealth:
|
||||
items:
|
||||
$ref: '#/definitions/httpapi.AdminRouteHealth'
|
||||
type: array
|
||||
type: object
|
||||
httpapi.AdminExecutionPoolWorker:
|
||||
properties:
|
||||
activeTasks:
|
||||
type: integer
|
||||
allocated:
|
||||
type: integer
|
||||
capabilities:
|
||||
additionalProperties: {}
|
||||
type: object
|
||||
heartbeatAt:
|
||||
type: string
|
||||
instanceId:
|
||||
type: string
|
||||
pressureState:
|
||||
type: string
|
||||
protocolVersion:
|
||||
type: string
|
||||
revision:
|
||||
type: string
|
||||
safeCapacity:
|
||||
type: integer
|
||||
workerId:
|
||||
type: string
|
||||
type: object
|
||||
httpapi.AdminRouteHealth:
|
||||
properties:
|
||||
connectTlsP95Ms:
|
||||
type: integer
|
||||
consecutiveFailures:
|
||||
type: integer
|
||||
consecutiveSuccesses:
|
||||
type: integer
|
||||
expiresAt:
|
||||
type: string
|
||||
firstByteP95Ms:
|
||||
type: integer
|
||||
jitterP95Ms:
|
||||
type: integer
|
||||
poolId:
|
||||
type: string
|
||||
routeProfileKey:
|
||||
type: string
|
||||
sampleCount:
|
||||
type: integer
|
||||
sampledAt:
|
||||
type: string
|
||||
state:
|
||||
type: string
|
||||
successRate:
|
||||
type: number
|
||||
uploadBytesPerSecond:
|
||||
type: number
|
||||
type: object
|
||||
httpapi.AdminTaskListResponse:
|
||||
properties:
|
||||
items:
|
||||
@@ -4449,6 +4562,8 @@ definitions:
|
||||
type: string
|
||||
podUid:
|
||||
type: string
|
||||
poolId:
|
||||
type: string
|
||||
preparingTasks:
|
||||
type: integer
|
||||
pressureReason:
|
||||
@@ -5816,6 +5931,24 @@ paths:
|
||||
summary: 重试计费结算
|
||||
tags:
|
||||
- billing
|
||||
/api/admin/runtime/execution-pools:
|
||||
get:
|
||||
produces:
|
||||
- application/json
|
||||
responses:
|
||||
"200":
|
||||
description: OK
|
||||
schema:
|
||||
$ref: '#/definitions/httpapi.AdminExecutionPoolResponse'
|
||||
"500":
|
||||
description: Internal Server Error
|
||||
schema:
|
||||
$ref: '#/definitions/httpapi.ErrorEnvelope'
|
||||
security:
|
||||
- BearerAuth: []
|
||||
summary: 查询逻辑执行池、Worker 容量和上游路由健康状态
|
||||
tags:
|
||||
- Admin Runtime
|
||||
/api/admin/runtime/model-rate-limits:
|
||||
get:
|
||||
description: 管理端查看平台模型维度的限流和冷却状态。
|
||||
|
||||
Reference in New Issue
Block a user