Rebooted the two oldest EC2 instances in our network today. It went as badly as expected. They were supposed to be identical

* One server rebooted perfectly
* Second server rebooted with data loss, permission issues, configuration failures.

How did a reboot cause data loss?

Lots of people dun goofed

@molaram doesn't have to be that. Could be that it had corruptions and for example the fs check made out worse or some other boot task (mysql migration/recovery for example) destroyed data. Mount points wrongly configured can assist with this too.
Could be that disk had to be restored to a backup due to a boot issue.

@Root is right. Messed up on so many levels. I don't think the initial team knew these instances were using ephemeral storage.