[yellowapple]

Something similar happened to me on IBM Cloud, back when it was called Bluemix. This was early in my career for an early-stage startup (that eventually shuttered - not due to this issue, but rather due to a lack of investor interest and us running out of runway as a result). Easy enough setup: staging and production, each with two Docker containers, one for the (Elixir/Sugar) app and one with the (Postgres) database.

Well, one day the production Postgres container just... vanishes. All the storage is gone. After weeks back and forth with IBM's support, they confirm that the loss is permanent. No explanation, no refunds, just "lol fuck you". Naturally, we didn't have backups yet, so the few users we did have now had to start from scratch.

I, too, learned my lesson about not doing an offsite backup regularly - and the bigger lesson that Bluemix should've been a last resort option. Funny enough, we had migrated to that from Heroku; we ended up migrating again to AWS (specifically: Elastic Beanstalk).