fix(similarasin): LLM 切换 gemini-3.8-flash + 修复双实例 stale 判死盲区

- 模型切换:glm-5.3-flash / gemini-3.7-flash(类目/合规/图片对比,appearancepatent)→ gemini-3.8-flash
  涉及 SimilarAsinProperties、AppearancePatentProperties、application.yml、TLS 探针与测试断言
- stale 判死盲区:finalizeStaleTasks 去掉 isOwnerCurrent 过滤,改由 job 锁持有实例全局判死
  (任务 27117 线上死因:客户端 done 信号被任务锁拒后断线,owner 过滤导致无人判死,
  被另一实例 job stuck 扫描误判为"文件生成任务运行超时";appearancepatent 同源盲区一并修复)
- 客户端回传重试:3x2s → 8 次,40902 锁忙固定等 15s(app_client/similar_asin.py,随客户端发包生效)

backend-java/src/test/java/com/nanri/aiimage/modules/appearancepatent/client/AppearancePatentLlmClientHttpTest.java

@@ -126,7 +126,7 @@ class AppearancePatentLlmClientHttpTest {
         assertThat(result.getAppearanceReason()).contains("视觉拆解");
         assertThat(result.getTitleReason()).contains("Apple");
         assertThat(callCounts.get("deepseek-v4-flash")).hasValue(1);
-        assertThat(callCounts.get("gemini-3.7-flash")).hasValue(1);
+        assertThat(callCounts.get("gemini-3.8-flash")).hasValue(1);
     }
 
     @Test
@@ -162,7 +162,7 @@ class AppearancePatentLlmClientHttpTest {
         AppearancePatentResultRowDto result = rows.get(0);
         assertThat(result.getAppearanceRisk()).isEqualTo("外观识别异常");
         assertThat(callCounts.get("deepseek-v4-flash")).isNull();
-        assertThat(callCounts.get("gemini-3.7-flash")).isNull();
+        assertThat(callCounts.get("gemini-3.8-flash")).isNull();
     }
 
     @Test
@@ -182,7 +182,7 @@ class AppearancePatentLlmClientHttpTest {
         client.inspectRows(List.of(row("6", "B006", "普通数据线", "", "https://img.example.com/6.jpg")), null, "test-key");
 
         String appearanceBody = capturedBodies.stream()
-                .filter(b -> b.contains("gemini-3.7-flash"))
+                .filter(b -> b.contains("gemini-3.8-flash"))
                 .findFirst()
                 .orElseThrow();
         assertThat(appearanceBody).contains("https://img.example.com/6.jpg");

backend-java/src/test/java/com/nanri/aiimage/modules/similarasin/service/LlmGatewayTlsProbe.java

@@ -61,7 +61,7 @@ public class LlmGatewayTlsProbe {
         }
         HttpClient client = builder.build();
         try {
-            String body = "{\"model\":\"glm-5.3-flash\",\"messages\":[{\"role\":\"user\",\"content\":\"ping\"}],\"max_tokens\":10}";
+            String body = "{\"model\":\"gemini-3.8-flash\",\"messages\":[{\"role\":\"user\",\"content\":\"ping\"}],\"max_tokens\":10}";
             HttpRequest request = HttpRequest.newBuilder()
                     .uri(URI.create("https://" + host + "/v1/chat/completions"))
                     .timeout(Duration.ofSeconds(30))