問題文
A team raises the maximum delivery count from 10 to 200 so that transient downstream outages no longer send messages to the dead-letter queue. What new problem does this create?
選択肢
- The queue's time to live is automatically extended so that the raised delivery count can be used in full, which is why the change appears to work
- Messages lose their session identifier after the first ten deliveries, so ordered processing silently breaks for anything that is retried more than ten times
- Messages that fail for a permanent reason are retried far longer, so the queue's backlog grows and genuinely broken items take much longer to surface
- The dead-letter queue stops accepting messages once the delivery count limit is raised above a supported threshold, so permanently failing messages are dropped instead of retained