Bar chart: cheapest new HA PostgreSQL cluster per month, DigitalOcean after 30 November 2026 $260, today $60.90, tport.cloud $60

A managed PostgreSQL database is the one piece of infrastructure most small teams happily pay extra for: someone else does the backups, the failover and the upgrades. DigitalOcean's managed PostgreSQL has been a popular choice because a standby was cheap. That changes this autumn.

The change

From DigitalOcean's PostgreSQL pricing docs: "Beginning 15 October 2026 for new accounts and 30 November 2026 for all accounts, new Standard Edition clusters cannot use plans with more than 4 GiB of RAM or add standby or read-only nodes." Existing clusters keep working as they are. For anything new that needs a standby (automatic failover when the primary dies) the path is Advanced Edition, which starts at $130 a month for an 8 GiB / 2 vCPU primary, and the standby bills at the same rate. The cheapest new high-availability cluster there becomes $260 a month.

Side by side

DigitalOcean today (Standard)DigitalOcean, new clusters from 30 Nov 2026tport.cloud
Single node$30.45: 1 vCPU · 2 GiB · 30 GiBsame (4 GiB or less only)$30: 6 vCPU · 12 GB · 100 GB NVMe
12-16 GB single node$244.35 (16 GiB · 6 vCPU)Advanced only, from $130 (8 GiB)$30
High availability$60.90: 2 GiB primary + standbyAdvanced: from $260 (8 GiB each)$60: 6 vCPU · 12 GB each
Backupsdaily, point-in-time restore 7 dayssamedaily, point-in-time restore 7 days
Connection poolerPgBouncer poolssamePgBouncer included (port 6432)

For reference, DigitalOcean's Standard single-node sizes today: 1 GiB $15.15, 2 GiB $30.45, 4 GiB $60.90, 8 GiB $122.10, 16 GiB $244.35, with a standby costing the same again. Extra storage is $0.215 per GiB a month on Standard. Our single node is one size, 12 GB of RAM and 100 GB of NVMe, because at $30 there is no reason to sell less.

tport.cloud console, Create a PostgreSQL cluster: $30 a month single node with 6 vCPU, 12 GB RAM, 100 GB NVMe, or $60 high availability with a primary and hot standby on a second server and automatic failover; region Europe or United States; PostgreSQL 18; pooler on 6432
Creating a cluster in the tport.cloud console: pick single node or HA, region, name.

What you get with ours

PostgreSQL 18 on its own server (two servers for HA), a role and a database ready, TLS required, direct connections on 5432 and PgBouncer on 6432. Nothing is open to the internet until you add your own IP addresses. HA runs a primary and a hot standby with automatic failover, and the connection string lists both hosts with target_session_attrs=read-write, so standard PostgreSQL drivers always find the primary. Daily backups plus WAL give point-in-time restore for 7 days, the same window DigitalOcean offers.

Who should move

If you already run a Standard cluster with a standby on DigitalOcean, nothing changes for it today. The moment it matters is your next new cluster, a new region, or a bigger plan: from 30 November that is Advanced Edition pricing. If you are about to create one, compare the $260 with our $60 first. Moving a small database is a pg_dump and a pg_restore with the two connection strings.

Sources

DigitalOcean: managed database pricing, PostgreSQL pricing details (the Standard Edition change and Advanced Edition prices) and PostgreSQL features (7-day point-in-time restore), checked 30 October 2026. tport.cloud: tport.cloud/postgresql. Our prices are the Germany prices from the pricing page; other regions cost a little more and the order form shows the exact price before you pay.

A PostgreSQL 18 cluster with 6 vCPU and 12 GB of RAM for $30 a month, or primary + standby for $60.

Managed PostgreSQL