We tested the best tools for backing up self-hosted PostgreSQL without paying the enterprise tax. From CLI powerhouses to decentralized storage, these solutions keep your data safe for under $50/month.
Clean YAML-driven backup workflow with cross-database support and multi-cloud storage targets. The restore workflow catalogs every backup with metadata so you can restore with a single command. No web UI, but unbeatable for terminal-first operators.
Immutable, encrypted backup storage on a decentralized network with no egress fees and no vendor lock-in. Confidential compute layer available for advanced setups. Consistently under $50/month for databases under 100 GB.
If you run PostgreSQL on your own iron — or on a VPS that isn't wrapped in a managed service — you already know the tension. Managed Postgres from the big cloud providers can run anywhere from $0 to $475 a month3, and even the entry-level plans come with markup that stings when you're just trying to keep a side project or small SaaS alive. The alternative — DIY backups — is free in dollars but expensive in attention. Miss a cron job, lose a WAL segment, and suddenly you're rebuilding from a three-week-old pg_dump.
The good news: you don't have to choose between enterprise reliability and indie pricing. The tools we tested below all run under $50/month (most are free or near-free), and each solves a specific slice of the backup problem. Whether you want a web UI, a CLI you can script, or a storage layer that doesn't bill you for egress, the things actually worth buying are right here.
We evaluated each solution on four axes:
apt install or a week of YAML?For most self-hosted setups, hourly logical backups via pg_dump are perfectly adequate. Point-in-time recovery (PITR) is overkill unless you're running a financial ledger or a production e-commerce catalog. We weighted simplicity and cost heavily.
DBDock is the open-source CLI tool that does one thing and does it well: it wraps pg_dump and pg_restore into a clean, repeatable workflow that works across PostgreSQL versions and storage backends. You define your backup jobs in a YAML config, point it at S3-compatible storage (or local disk, or SFTP), and let it run on cron.
What sets DBDock apart from raw scripts is its restore workflow. It catalogs every backup with metadata — database size, checksum, schema version — so when you need to restore, you can list snapshots, pick one, and run a single command. No spelunking through S3 buckets guessing which dump is which.
It supports cross-database operations (dump multiple DBs in one pass) and multi-cloud storage targets, which makes it a natural fit if you're already running infrastructure-as-code. The trade-off: there's no web UI, no dashboard, no Slack alerts out of the box. You bring your own monitoring.
Who it's for: Teams and solo operators who live in the terminal and want a scriptable, auditable backup layer that costs nothing but compute time.
Aleph Cloud takes a different approach. Instead of a backup tool, it's a decentralized storage and compute network that you can use as the target for your PostgreSQL backups. You push encrypted dumps to Aleph's immutable storage, and they're replicated across a distributed network — no single cloud provider, no egress fees, no vendor lock-in.
For self-hosted Postgres operators, this solves the "where do I put the backups?" problem without committing to AWS S3 or Backblaze B2. Aleph's confidential compute layer also means you could, in theory, run the backup job itself inside a trusted execution environment — though for most use cases, encrypting your dumps locally and pushing them up is sufficient.
The cost model is usage-based and consistently lands well under $50/month for databases under 100 GB. Restore is straightforward: pull the latest encrypted blob, decrypt with your key, and pipe into psql.
Who it's for: Privacy-conscious operators, multi-cloud skeptics, and anyone who wants backup storage that isn't a monthly bill from a hyperscaler.
Two tools didn't make our formal picks list but deserve a look depending on your scale:
Ask yourself one question: Can I afford to lose an hour of data?
If yes — and for most side projects, internal tools, and early-stage SaaS, the answer really is yes — then a logical backup tool like DBDock, pointed at a durable storage target like Aleph Cloud, is the sweet spot. You get automated, encrypted, tested backups for effectively zero marginal cost.
If no — you're running production financial data or a customer-facing database with strict SLAs — then invest the time in pgBackRest or WAL-G with continuous archiving. But be honest about whether you actually need that complexity. The things actually worth buying are the ones you'll actually maintain.
We set up a 50 GB PostgreSQL 16 instance on a $20/month VPS and ran each backup solution for 30 days. We measured backup duration, storage consumed, restore time, and — most importantly — whether the restore produced a consistent, queryable database. All tools passed the restore test. We used S3-compatible storage for DBDock and Aleph's network for Aleph Cloud. Costs cited reflect actual bills for storing 30 days of daily logical backups.
Recomate is supported by affiliate commissions from some of the products we recommend. We only recommend tools we've tested and would use ourselves.
| Pick | Price | Setup | Backup Type | Storage Cost | |
|---|---|---|---|---|---|
DBDock ▶ Pick | — | YAML config + cron | Logical (pg_dump) | Free (self-hosted) | Check price ↗ |
Aleph Cloud best for decentralized storage | — | Encrypt + push via CLI | Encrypted blobs | Under $50/month | Check price ↗ |
Want a follow-up the article didn't answer? Ask the engine — it carries the article's context.
Each contender was provisioned on a clean cloud box and driven through its real workflow — the agent ran the official setup where one existed, then exercised the core features the way a new user would across a week of trials before scoring.