1. 检查alertmanager配置文件中的resolve_timeout参数，确保它正确设置。例如：

route:
  group_wait: 5s
  group_interval: 5s
  repeat_interval: 5m
  receiver: default-receiver
  routes:
  - match:
      alertname: Watchdog
    receiver: ops-team
  - match_re:
      severity: ^(warning|critical)$
    receiver: ops-team
    routes:
    - match:
        product: foo
      receiver: foo-team
    - match:
        product: bar
        severity: critical
      receiver: bar-team
 
receivers:
- name: default-receiver
  webhook_configs:
  - url: 'http://localhost:5001/alert'
    send_resolved: true
    http_config:
      timeout: 30s
      bearer_token_file: /etc/token
    ### 默认情况下，send_resolved是false，这意味着它不会发送resolve通知。确保将其设置为true。 

- name: ops-team
  webhook_configs:
  - url: 'http://localhost:5001/alert'
    send_resolved: true

- name: foo-team
  webhook_configs:
  - url: 'http://localhost:5001/alert'
    send_resolved: true

- name: bar-team
  webhook_configs:
  - url: 'http://localhost:5001/alert'
    send_resolved: true

route:
  group_by: ['alertname']
  group_wait: 10s
  group_interval: 10s
  repeat_interval: 1h
  receiver: webhook

2. 检查接收器配置，确保在触发警报和解决警报时都设置了相应的send_resolved值。例如，上面的示例中，所有接收器都设置了send_resolved为true，这意味着在警报解决时也会发送通知。

3. 检查警报接口是否配置正确。如果警报接收器使用Webhook，则确保Webhook上的URL是正确的，并确保在接收警报的端点上实现了支持解决通知。

4. 检查警报的解决状态是否被正确标记。在Prometheus中，警报的解决状态是通过标记警报的标签之一来识别的。如果标记的值不正确，则可能会导致Alertmanager无法识别警报的