Downtime this morning was super annoying, but no data loss and everything should be running smoothly again. Brainstormed with @vincent ideas to make it more robust. We have redundancy in photo storage, but when the primary storage is down, it cascades to other errors. Fixable.