Datastores

A datastore is a managed database attached to a service environment. Kuberns supports PostgreSQL, MySQL, and MongoDB.

You create a datastore from an environment's Resources tab. Once it exists, it also appears under Datastores in the platform navigation, where it gets its own pages. The listing shows each database with its environment, type, status, and creation date.

Overview

The Overview page holds the connection details your application needs — database name, username, password, and hostname — along with current memory and storage usage and the latest backup information.

Keep application-facing credentials in environment variables rather than in your repository.

Metrics

The Metrics page reports RAM and storage usage over time, and is where alert configuration for the datastore is available.

Backups

You can take a backup at any time, and you can enable automatic backups on a fixed interval. Completed backups are listed with their status, and each one can be downloaded or deleted.

Restoring a database

Kuberns can restore a backup directly into a datastore for you. You can restore from either:

• An existing Kuberns backup — select Restore next to any completed backup in the backup history.
• Your own dump file — upload it and restore from that instead.

You can start a restore from three places:

• The datastore's Backups page.
• The Resources tab of the environment, using Restore Backup on the database resource. The database must be running to use this.
• The Add Resource flow, by attaching a backup file to restore into a new database once it's created.

Uploading your own file

Accepted file types depend on the database engine:

Engine	Accepted extensions
PostgreSQL	.sql, .sql.gz, .dump
MySQL	.sql, .sql.gz
MongoDB	.archive, .archive.gz, .gz

Uploads are capped at 5 GB.

What happens during a restore

Restoring replaces all current data in the datastore. Before restoring into a datastore that already holds data, Kuberns automatically takes a safety backup first, so the prior state can be recovered if needed. This safety-backup step is skipped when restoring into a brand-new, empty database.

Only one restore can run at a time per datastore — a new one can't be started while another is in progress. Track progress from the datastore's Backups page, which shows restore status alongside backup history.

If a restore fails partway through, Kuberns does not automatically roll back to the safety backup — restore from it manually if needed.

Settings

Settings exposes the database password, storage upgrades for the current plan, IP whitelisting, and deletion of the datastore.

Deleting a datastore is irreversible. Download any backup you still need first.

IP whitelisting

By default, a datastore accepts connections from any IP address, secured only by its password. IP whitelisting restricts the database's port to a list of IP addresses or CIDR ranges you choose.

Turn it on from the datastore's Settings tab — the same control is also available inline on the database's row under the environment's Resources tab. The datastore must be running to change this setting; while it's stopped, you can still view the current configuration.

Servers and background workers in the same environment are always allowed through, on or off — you don't need to add them yourself. Add an entry for any other client that needs to connect, such as your own machine, an office network, or a CI runner. There's no automatic detection of your current IP, so add it explicitly before turning whitelisting on, or you could lose access from that client.

Each entry must be:

• IPv4.
• A public (non-private) address.
• A prefix of /16 or narrower — very wide ranges, including one that would reopen the database to everyone, aren't accepted as an entry.

You can add up to 30 entries.

Turning whitelisting off immediately reopens the database to all IPs and clears the saved list. Turning it back on later starts from an empty list, not the one you had before.