优化架构K8s 副本和队列

k8s/core/api-deployment.yaml

@@ -5,7 +5,7 @@ metadata:
   labels:
     app: airshelf-core-api
 spec:
-  replicas: 1
+  replicas: 2
   selector:
     matchLabels:
       app: airshelf-core-api

k8s/core/worker-deployment.yaml

@@ -23,7 +23,8 @@ spec:
         # Celery worker connects to the external (Volcano managed) Redis broker
         # configured via the airshelf-core-env secret. Uses `args` (not `command`)
         # so the image ENTRYPOINT still runs but skips migrate/collectstatic ($1=celery).
-        args: ["celery", "-A", "airshelf.celery:app", "worker", "-l", "info", "-Q", "celery,airshelf.quick", "--concurrency", "4"]
+        # 本 Pod 只吃默认 celery 队列(出图/出片等慢活)。提取/通知/脚本思考在 worker-quick。
+        args: ["celery", "-A", "airshelf.celery:app", "worker", "-l", "info", "-Q", "celery", "--concurrency", "4"]
         envFrom:
         - secretRef:
             name: airshelf-core-env

k8s/core/worker-quick-deployment.yaml

@@ -0,0 +1,42 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: airshelf-core-worker-quick
+  labels:
+    app: airshelf-core-worker-quick
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: airshelf-core-worker-quick
+  template:
+    metadata:
+      labels:
+        app: airshelf-core-worker-quick
+    spec:
+      imagePullSecrets:
+      - name: cr-pull-secret
+      containers:
+      - name: airshelf-core-worker-quick
+        image: ${CI_REGISTRY_IMAGE}/airshelf-core-api:latest
+        imagePullPolicy: Always
+        # 只消费 airshelf.quick:提取角色、通知补齐、脚本思考、视频状态轻轮询。
+        # 与出片 worker 隔离,Seedance 占满时提取仍能跑。
+        args: ["celery", "-A", "airshelf.celery:app", "worker", "-l", "info", "-Q", "airshelf.quick", "--concurrency", "4"]
+        envFrom:
+        - secretRef:
+            name: airshelf-core-env
+        livenessProbe:
+          exec:
+            command: ["sh", "-c", "celery -A airshelf.celery:app inspect ping"]
+          initialDelaySeconds: 40
+          periodSeconds: 60
+          timeoutSeconds: 15
+          failureThreshold: 3
+        resources:
+          requests:
+            memory: "256Mi"
+            cpu: "100m"
+          limits:
+            memory: "1Gi"
+            cpu: "500m"