BasePilot
Sign inCreate account

Managed PostgreSQL on Hetzner and UpCloud

BasePilot provisions and operates PostgreSQL clusters on the European cloud providers Hetzner and UpCloud, in your own cloud account or on your own servers. Replication, failover, continuous backups and major version upgrades are handled for you, from Germany.

Operated from Germany. Billed per node and hour. No minimum term.

Example cluster

shop-x7k2f

PostgreSQL 17

Running

Connection string

postgresql://app@shop-x7k2f.postgres.basepilot.de:5432/defaultdb?sslmode=verify-full

Nodes

  • n1.shop-x7k2f.postgres.basepilot.dePrimaryHealthy
  • n2.shop-x7k2f.postgres.basepilot.deReplicaHealthy
  • n3.shop-x7k2f.postgres.basepilot.deReplicaHealthy

PostgreSQL at a glance

What a PostgreSQL cluster on BasePilot consists of.

Versions
17, 16, 15
Default version
PostgreSQL 17
Port
5432
Nodes per cluster
1 to 3
Replication
Streaming replication, asynchronous
Backups
Continuous WAL archive and scheduled base backups, encrypted before upload
Point-in-time restore
Yes
Major version upgrades
Yes, blue-green
Availability
Generally available

What managed PostgreSQL includes

Specific to PostgreSQL. Monitoring, maintenance windows, access control and hourly billing come with every cluster.

  • Streaming replication and automatic failover

    Run up to three nodes; replicas follow the primary through PostgreSQL's streaming replication. If the primary fails, BasePilot isolates it, promotes the replica that has received the most data and rebuilds the failed node. The cluster's address stays the same.

  • Point-in-time restore

    The write-ahead log is archived to object storage continuously and base backups run on your schedule. Restore into a new cluster from a backup, or to any point in time the archive covers.

  • Encrypted backups

    Backups and archived log segments are encrypted on the node before they are uploaded, so no table data reaches object storage unencrypted.

  • Major upgrades without a long outage

    A blue-green upgrade builds a second cluster on the new major version, keeps it in sync through logical replication and switches over when you are ready.

  • Databases, users and access

    Create databases and users in the console. Users are read-only, read-write or admin and can be limited to individual databases.

  • TLS and SCRAM only

    Remote connections require TLS and SCRAM authentication. Every cluster has its own certificate authority, so clients connect with sslmode=verify-full; alternatively, you upload your own certificate.

Everything that comes with every cluster

Built and operated in Europe

BasePilot is a German service on European infrastructure. You decide in which data centre each cluster runs.

Operator in Germany
BasePilot is developed and operated in Erfurt, Germany. Contracts are governed by German law.
European cloud providers
The servers come from Hetzner (Germany) and UpCloud (Finland). You choose the data centre location for every cluster yourself, for example in Germany or Finland.
Data processing under the GDPR
We process the data in your databases solely on your behalf and on your instructions. A data processing agreement under Art. 28 GDPR is available on request.
Analytics only with consent
This website measures visits only if you agree; without your consent it sets no cookies. The console uses only technically necessary cookies.

Questions about managed PostgreSQL

Short answers on versions, operation and limits.

Which PostgreSQL versions does BasePilot offer?
PostgreSQL 17, 16, 15. New clusters use PostgreSQL 17 by default.
How does failover work for PostgreSQL?
BasePilot monitors the primary of a multi-node cluster continuously. When it stops answering and a second check confirms the failure, the old primary is isolated, the most advanced replica is promoted and the failed node is rebuilt as a replica. This usually takes a minute or two. Replication is asynchronous, so transactions that had not yet reached the replica can be lost.
How far back can I restore?
To any point in time since your oldest retained backup. You set how often backups run and how many are kept. A restore always creates a new cluster and leaves the original untouched.
What does a major version upgrade require?
Every table needs a primary key, because the upgrade uses logical replication. After the switch the old cluster remains available read-only until you delete it.

Start with your first cluster

Create an account, choose an engine, a size and a region, and connect within minutes.