Backups and plans
Take on-demand backups, understand their retention and restore path, and add or retire database slots without downtime.
Managed databases include on-demand backups. There is no cron editor and no bucket configuration: you ask for a backup, the platform takes it, and the console keeps the history.
Take a backup
- Open the database.
- Choose Create backup.
- Watch the Backups table: the row moves from pending to complete.
A backup captures the database at that moment. Do it before schema changes, bulk imports, engine upgrades or anything else you might want to undo.
Retention and restore
The table lists Status and Created for each backup. It does not expose a restore, download or delete action today.
To restore, open a support ticket with:
- The database name and engine
- The backup timestamp you want
- Where the restore should go (a new database, or over the current one)
Support confirms before overwriting a live database. Restoring over a live instance replaces its contents, so point the application at a staging copy first when the data matters.
Backups are not a substitute for application-level exports. If you need a portable dump you control, run one from the machine that can reach the database:
pg_dump "$DATABASE_URL" --format=custom --file=backup.dump
mongodump --uri "$MONGO_URL" --archive=backup.archive --gzip
Then copy the dump to a Z2 bucket. That gives you a copy independent of the platform.
Slots and plans
- One database instance = one slot.
- Plans add up. Buy another plan when the current ones are full; slots combine.
- Per-database resources come from the plan the slot was taken from. A database created from a Starter slot has 0.5 vCPU and 512 MB, even if you also own a Scale plan.
- Retiring a plan with databases on it is handled by support: the databases are moved first, then the plan is closed.
The Database plan balance card on the Databases page shows used and total slots, plus how many are free.
Changing resources
There is no resize control in the console. Options:
- Add a larger plan, then create new databases from its slots. Existing instances keep their original resources.
- Migrate the database: dump, create a new instance on the larger plan, restore, switch the app's connection string, then retire the old one through support.
For a growing production database, plan 2 is the usual path. Do it during a maintenance window and keep the old instance until the new one is verified.
Engine upgrades
Version upgrades are not self-service. Open a ticket with the engine, current version and target version. Support takes a backup first and schedules the work.
What is not included
- No SSH or root access to the database host.
- No custom extensions unless they ship with the stock engine image. Ask support before depending on one.
- No replica or failover controls in the console.
- No backup schedule editor. Backups are on demand.
If you need replicas, custom configuration or a specific extension, a Cloud VPS running the database yourself is the honest alternative.
Next
- database backup
- restore
- slots
- plan
- scaling
- retention
Ready to use this
PostgreSQL, MySQL, MongoDB or Redis on managed infrastructure.
- Any engine from the same slot pack
- Deploying from a slot costs nothing extra

