Managed databases
Letting CrudAdmin create real databases on a tenant’s behalf, instead of the tenant only registering ones it already has.
Table of contents
The provisioning account
A platform administrator configures one credential the whole platform uses to actually create databases, in Host Settings — one of three kinds:
| Kind | What it needs |
|---|---|
| A MySQL server you run | An admin login that can CREATE/CREATE USER/GRANT. |
| Plesk server | Panel address, target subscription, and an admin login or secret key. |
| cPanel account | Panel address, account username, and an API token. |
Per-tenant mode
Configured per tenant from Admin console → Tenants → Databases:
| Mode | Behavior |
|---|---|
| Off | The tenant only registers databases it already has elsewhere. |
| Ask first | The tenant requests a database by name; a platform administrator approves or rejects each request. |
| Self-service | The tenant creates its own managed databases directly, up to a limit the platform administrator sets. |
Both the “ask first” and “self-service” modes also set a name prefix — how the database is actually named on the underlying server, which is never exposed in the tenant’s own API calls (those always address a database by its own db_key, independent of the real server-side name).
Requesting or creating one
From the tenant side, this surfaces on the Databases tab — either a request form (ask-first mode) or a direct “create a managed database” action (self-service mode), both counted against whatever overall database limit the platform administrator has set for that tenant.
What happens on approval
Approving a request (or a self-service creation) genuinely calls out to the real provisioning account — Plesk, cPanel, or a MySQL server — which can take a few seconds, not instant. A request that was approved but failed to actually provision (a name collision, an unreachable account) stays visible with its error and a retry option, rather than silently disappearing.