3 comments

[ 5.3 ms ] story [ 19.4 ms ] thread
At least their backups worked! Still up to human error though. Hard to get around human error.
No, it's not that hard. It's more about getting the priorities right.

The developer and production environments should be separated at a deep network level making it very very difficult for anything (rogue script, rogue admin, etc) to access production environment accidentally.

DigitalOcean are good people for sharing this. They provide a good service and good price point.

It's funny how this outage was similar to Amazon's a month or so ago. Automated playbooks with high user credentials have been a bit of a problem lately!!