I accidentally destroyed a production Managed PostgreSQL cluster today
I was working with a separate fork/test database and mistakenly triggered the destructive action on the production cluster from the DigitalOcean control panel. It was completely accidental.
I understand that DigitalOcean's documentation says that destroying a Managed Database cluster also destroys its automated backups.
However, I'm hoping someone from DigitalOcean or someone with experience with their infrastructure can clarify whether there is any possible recovery path after cluster destruction.
Specifically, I'm wondering whether DigitalOcean may have any backend/infrastructure-level copy that isn't exposed through the normal dashboard/API
The cluster contained important production data. Unfortunately, our most recent independent PostgreSQL dump is approximately 3 days old, so we potentially have around 3 days of data that isn't in that dump.
I've already opened an urgent DigitalOcean support ticket and requested that it be escalated to the database/infrastructure engineering team, also .
If anyone from DigitalOcean has dealt with an accidentally destroyed Managed Database cluster before, I'd really appreciate knowing what recovery options were available.
This is a production incident and extremely time-sensitive. Thank you.
---
UPDATE: they reached back to me quickly , they are "working on it".
---
UPDATE: Less than an hour later, they came back with the news:
The production PostgreSQL cluster had been restored and was back online.
Huge thanks to the DigitalOcean support and engineering teams. The speed and effort were honestly incredible. they got it up in just 1hr which is incredible. This experience has given me even more confidence in DigitalOcean and its services. I can wholeheartedly vouch for DigitalOcean and the quality of its support.
i never imagined this could happen. you could be the biggest fool sometimes and delete something without double checking.. i don't know it was sleep or stress but the incident exposed several flaws in my own disaster-recovery design. never back down form spending extra on multiple local backups. i use to see memes on this but hell this is going to be a lifetime trauma, the last 1hr was absolute nightmare.