[Question] Is the logic of PING connectivity reasonable?

### Question

### 当前逻辑
ping 连通时返回ping耗时，未连通、异常时没有返回ping耗时。[相关代码](https://github.com/dromara/hertzbeat/blob/master/collector/src/main/java/com/usthe/collector/collect/icmp/IcmpCollectImpl.java)
### 问题
ping告警只能根据ping耗时触发，ping采集失败和异常情况没有ping 耗时。不能触发告警，而发现ping异常、超时正是ping检测的目的。是比ping耗时变长更严重的问题
### 建议
#### 方案一
ping采集发生异常、超时将ping耗时赋值为Long最大值，表示无穷大，可触发告警
#### 方案二
将采集失败、超时等异常情况添加到PING连通性的告警表达式配置中，可配置发生异常时告警

Start by reading collector/src/main/java/com/usthe/collector/collect/icmp/IcmpCollectImpl.java and trace how successful, failed, and timed-out PING results reach alert expressions. Compare the two proposed approaches and confirm that the chosen behavior lets failed or timed-out checks trigger alerts without losing normal ping-latency alerts.

Written by the indexing model from the issue text.