perf(queue): 限制大媒体任务内存并消除准入惊群

将同步与异步非文本任务的准入唤醒改为按任务和全局队首推进，批量续租等待者，避免大量请求同时争抢 advisory lock 和数据库连接。

对 Base64 请求解析、素材解码、上游媒体执行和结果物化增加分层并发限制，复用请求素材并释放重复 Gemini wire 数据；生产 API 默认入口解析并发 16，媒体物化并发 8。

新增 1000 个同步 Gemini 图像编辑请求的模拟上游压力验收，10 MiB 输入和输出下全部成功，Heap 峰值增长约 2.58 GiB，并验证共享上传哈希、单次 attempt 与本地零落盘。

验证：go test ./... -count=1；go vet ./...；gofmt -l 无输出；kubectl kustomize deploy/kubernetes/production；10 MiB Gemini Base64 千任务压力测试通过。

deploy/kubernetes/production/application.yaml

@@ -54,6 +54,10 @@ spec:
               value: "false"
             - name: AI_GATEWAY_DATABASE_MAX_CONNS
               value: "16"
+            - name: AI_GATEWAY_MEDIA_REQUEST_CONCURRENCY
+              value: "16"
+            - name: AI_GATEWAY_MEDIA_MATERIALIZATION_CONCURRENCY
+              value: "8"
             - name: POD_NAMESPACE
               valueFrom:
                 fieldRef:
@@ -179,6 +183,10 @@ spec:
               value: "false"
             - name: AI_GATEWAY_DATABASE_MAX_CONNS
               value: "16"
+            - name: AI_GATEWAY_MEDIA_REQUEST_CONCURRENCY
+              value: "16"
+            - name: AI_GATEWAY_MEDIA_MATERIALIZATION_CONCURRENCY
+              value: "8"
             - name: POD_NAMESPACE
               valueFrom:
                 fieldRef:
@@ -304,6 +312,8 @@ spec:
               value: "true"
             - name: AI_GATEWAY_DATABASE_MAX_CONNS
               value: "24"
+            - name: AI_GATEWAY_MEDIA_MATERIALIZATION_CONCURRENCY
+              value: "8"
             - name: POD_NAMESPACE
               valueFrom:
                 fieldRef:
@@ -429,6 +439,8 @@ spec:
               value: "true"
             - name: AI_GATEWAY_DATABASE_MAX_CONNS
               value: "24"
+            - name: AI_GATEWAY_MEDIA_MATERIALIZATION_CONCURRENCY
+              value: "8"
             - name: POD_NAMESPACE
               valueFrom:
                 fieldRef: