Hacker News Reader

Is this pure network issue (ie: strictly connectivity), or were instances rebooted?

Seems like network only. I have an instance that was unavailable via SSH. When things came back up I still had 81 days of uptime.