ruby-on-rails - EC2 instance running Rails app becomes unreachable

Question

I have a rails app running on several EC2 instances (using instance storage, not EBS). Frequently (more often than it seems like it should), an instance becomes unreachable (cannot ping or ssh, and Amazon ELB reports it as out of service). However, in the AWS console the instance still shows up as "running." Rebooting the instance doesn't help -- once an instance gets into this unreachable state, I can reboot it successfully, but it continues to be unreachable after reboot. The only alternative is to terminate it and start a new instance.

Any ideas why this might be occurring? Unfortunately I can't access the logs on the instance once it becomes unreachable, and my monitoring software does not show any spikes in any particular metric right before the instance becomes unreachable. The System Log for the instance that I can access through the AWS console is not informative (it seems to boot up fine).

Answer 1

没有看到服务器很难回答。但是，您可以通过停止实例的硬盘驱动器、从中卸载 EBS 驱动器并将其附加到活动实例来检查该实例的硬盘驱动器。然后，使用实时实例查看系统日志。我会查看硬盘驱动器是否已填满（这很容易占用服务器）。如果没有，我会查看系统日志以查看是否报告了错误。