Standalone, no site required
Buy a database without buying hosting. Your application can run anywhere; the database is a product in its own right.
Zinn® Managed Databases
Standalone managed databases: you do not need to host a website with us to run a database with us. PostgreSQL or MySQL, three tiers by size on each, with automatic failover from the medium tier upward. We verify restores by connecting to the restored database rather than trusting that a backup job finished.
Zinn® Managed Databases is PostgreSQL or MySQL that we run for you, sold on its own. Nothing about it assumes you also host a site here: point an application at it from wherever that application lives, and treat it as a database service rather than an add-on to a hosting plan.
There are three sizes on each engine. The small tier is a single instance. The medium and large tiers add automatic failover, so a lost node is a failover event rather than an outage you find out about from your users. What failover costs differs by engine — PostgreSQL arbitrates between two instances, while MySQL needs a three-member group and a pair of routers to reach a quorum — which is why the MySQL tiers are priced a little higher at every size.
Managed PostgreSQL and MySQL with the operational parts done properly, and stated plainly enough that you can plan around them.
Buy a database without buying hosting. Your application can run anywhere; the database is a product in its own right.
Medium and large run with automatic failover. The step up from small is a change in availability, not only a change in size.
The small tier runs one instance, which suits development, staging and lighter production workloads where a short recovery window is acceptable.
On PostgreSQL, recovery is to a moment rather than to last night's snapshot, which is what lets you undo a bad migration or a mistaken delete without losing the day around it. MySQL databases are backed up nightly and kept for fourteen nights.
A backup job reporting success is not evidence that a database can be restored. We verify restores by connecting to the restored database and reading from it.
Every tier runs the engine you chose. Moving up changes the vCPU, memory and disk allocation and, from medium, adds automatic failover.
Provision it, connect to it, and let us carry the operational half.
Choose PostgreSQL or MySQL, then small, medium or large by how much CPU, memory and disk you need and whether you want automatic failover.
You get credentials and a connection endpoint. The application can sit on our estate or on somebody else's; the database does not care.
We back up on a schedule and check that the backups restore by connecting to the result, so recoverability is measured rather than assumed. On PostgreSQL that schedule is continuous, so recovery is to a point in time.
Three tiers on each engine, billed monthly, sized by vCPU, memory and disk — with automatic failover included from the medium tier. Tier prices are computed from measured infrastructure unit rates and re-derived when those rates move.
$13.99/month
$13.99/month
$48.99/month
$73.99/month
$102.99/month
$153.99/month
Prices track measured infrastructure rates rather than being set once and left, so the tier you are on stays proportionate to what it actually costs to run.
Plenty of hosts will run a database for you, provided you also buy the hosting. This one is sold on its own, which means you can put your database with us and your application wherever it makes sense, without restructuring your stack to fit a bundle.
The two claims that matter most in a database product are the ones easiest to make loosely: that it fails over, and that it restores. Here, failover is a stated property of the medium and large tiers, and restores are verified by connecting to the restored database rather than by reading a green job log.
No. Managed Databases is standalone. Your application can run on your own servers, on another provider, or on our estate — the database is provisioned and billed independently.
Medium and large run with automatic failover. Small is a single instance, which is why the step from small to medium is an availability decision rather than only a capacity one.
Both engines are backed up and both are restore-tested by connecting to the result. They differ in how far back you can aim: PostgreSQL archives continuously, so you can recover to a moment — which matters most after a bad migration or an accidental delete, where you want everything up to the second before the mistake. MySQL is backed up nightly, with the last fourteen nights kept.
We restore them and connect to the result. A backup job reporting success only tells you a job finished; connecting to the restored database and reading from it tells you the data is there and usable. That is the check we run.
Yes. The tiers run the same engine you started on, so moving up changes the vCPU, memory and disk allocation and, from medium, adds automatic failover. Talk to us before a large migration and we will plan the move with you.
Pick an engine and a tier, connect your application, and get automatic failover from the medium tier upward — whether or not anything else of yours is hosted here.
Get started