[BUG] <title>pod异常退出后，节点没有正常下线

### Is there an existing issue for this?

- [x] I have searched the existing issues

### Current Behavior

环境：K8S容器部署
版本： 2.6.1
问题： pod节点因k8s node宕机异常退出后，服务出现访问异常，具体表现为有时很快，有时很慢（应该是超时之后转到的正常节点），查看admin页面，发现upsteam中异常节点没有下线，查看amin日志发现

`2025-05-12 01:16:17 [shenyu-upstream-health-check-request-6] ERROR org.apache.shenyu.admin.service.impl.UpstreamCheckService - check zombie upstream the url=10.244.47.9:8080 is fail

2025-05-12 01:16:30 [shenyu-upstream-health-check-request-3] ERROR org.apache.shenyu.common.utils.UpstreamCheckUtils - socket connect is error. host:10.244.47.9 port:8080 timeout:3000`

手动删除异常节点ip之后，服务正常

### Expected Behavior

_No response_

### Steps To Reproduce

_No response_

### Environment

```markdown
ShenYu version(s):
```

### Debug logs

_No response_

### Anything else?

_No response_

Start with the UpstreamCheckService and UpstreamCheckUtils entry points, using the reported zombie-upstream and socket-connect log messages as the first clues. Investigate the K8S node failure scenario described in the issue and verify that the failed upstream is removed without manual IP deletion, after which requests should no longer route to it.

Written by the indexing model from the issue text.