SimpleBackups vs pgBackRest: managed dumps or self-run point-in-time recovery

The core difference between SimpleBackups and pgBackRest

SimpleBackups vs pgBackRest is a managed service against an open source tool, and they back up PostgreSQL in different ways. SimpleBackups runs pg_dump on a schedule and stores the dump file. pgBackRest copies the database cluster's files and archives the write-ahead log (WAL), so you can replay changes to a chosen moment. That one difference drives recovery point, effort and cost. Facts here come from simplebackups.com, pgbackrest.org and the PostgreSQL manual, read on 3 October 2026.

Entry price for each

pgBackRest is free. Its site states it is open source under the MIT license with no restriction on commercial use. The current stable release is v2.59.2. You pay for the storage you point it at and for the engineer time to run it.

SimpleBackups has a free Basic plan limited to one job and a 1 GB database. Lite is $49 per month on monthly billing ($39 per month on yearly billing) with 5 jobs and databases up to 5 GB. Plus is $99 ($79 yearly) with 20 jobs and databases up to 250 GB. Details are on the SimpleBackups pricing page.

SimpleBackups and pgBackRest compared on nine criteria

Point-in-time recovery: what you gain with pgBackRest

The PostgreSQL manual explains the mechanism: the server keeps a write-ahead log of every change, and a file-level backup combined with archived WAL can be replayed to bring the database to a chosen state. The manual also says this approach is more complex to administer than a dump.

Take a table dropped by mistake at 14:10. With daily SimpleBackups dumps at 03:00 UTC, you restore the 03:00 file and lose about eleven hours of writes. With hourly dumps on Plus, you lose up to an hour. With pgBackRest you restore the last backup and replay WAL to 14:09. Dumps versus point-in-time recovery covers when that gap matters.

Operations burden: who does the work

The pgBackRest user guide walks through what you set up yourself:

• Install pgBackRest on the database host, and on a repository host if you use one.
• Write its configuration file: the cluster path, repository location, retention for full and differential backups, and optional repository encryption.
• Turn on WAL archiving in PostgreSQL by setting archive_mode and archive_command, then restart the server.
• Create the stanza and run pgBackRest's check command to confirm archiving works.
• Schedule full and incremental backups yourself.
• Watch for failed backups and stalled archiving. The PostgreSQL manual warns that while archiving fails, WAL files keep filling the server's pg_wal directory.
• Create object storage buckets in advance. pgBackRest does not create them.

With SimpleBackups the vendor handles scheduling, retries, alerts and anomaly checks. You fill in a form. Neither one tests restores for you. Managed versus scripted backups weighs that effort.

What each does that the other cannot, who should use which, and switching

pgBackRest only: recovery to a point between backups, and delta restore, which re-copies only files that differ from the backup. SimpleBackups only: backing up managed PostgreSQL where you have no shell on the host, plus MySQL, MongoDB and SaaS apps in the same dashboard.

• Use pgBackRest if you run PostgreSQL on your own servers, cannot lose an hour of writes, and have someone who will own the setup.
• Use SimpleBackups if your PostgreSQL is managed, your databases are small enough that dumps finish quickly, and losing up to one backup interval is acceptable.
• Use both if you want point-in-time recovery plus an independent logical copy that can be loaded into a different PostgreSQL server.

Switching. The two formats are not interchangeable. A pg_dump file cannot seed a pgBackRest repository, and a pgBackRest backup is not a dump. Moving to pgBackRest means the setup list above plus one server restart to enable archiving. Moving to SimpleBackups means creating a read-only database user and a backup job. In either direction, keep the old backups until the new system has a full retention cycle and you have restored from it once. The restore runbook gives the drill.