Resistro vs Databasus: Comparison

Factual comparison of the Resistro CLI and Databasus for PostgreSQL, MySQL and MariaDB backup. Measured where we ran it; vendor claims are marked. As of 25 October 2026.

Both tools are open source (Apache-2.0) and back up PostgreSQL and MySQL/MariaDB. Each cell carries its evidence level as a badge. Compared on 25 October 2026: the Resistro CLI from the public main branch, Databasus from its GitHub README. The Enterprise column is the self-hosted console (see Enterprise).

  • Run executed
  • Help read --help
  • Code code reading
  • Vendor Databasus docs
  • Docs Resistro docs

Comparison matrix

Characteristic Resistro CLI Resistro Enterprise Databasus
License Apache-2.0 Run Commercial licence, 14-day trial Docs Apache-2.0 Vendor
Form Single Go binary, no container Self-hosted console, web UI, agents on DB hosts Docs Server application with web UI Vendor
PostgreSQL: point-in-time recovery restore pitr to time, xid, lsn, name Help PITR activation and restore from console Docs Physical backups with WAL streaming Vendor
PostgreSQL incremental backups Yes backup cluster --physical --backup-type incremental (pg_basebackup --incremental), PostgreSQL 17+, wrapper engine Help No Full basebackup only Code Yes Physical incremental Vendor
MySQL / MariaDB: point-in-time recovery Yes Binlog archive and replay Help Yes Console PITR Docs No Logical backups only Vendor
Table-level point-in-time recovery No Yes Restore single tables from PITR window Docs No Vendor
MongoDB No No Yes Logical only Vendor
WAL capture archive_command, or wal stream --slot Code Agent archives WAL/binlog for PITR target Docs Continuous WAL streaming, parallel upload Vendor
Restore verification verify-restore and drill into isolated container Run Restore test and drill from console, auto-drill Docs Restore into container, row-count report Vendor
Prometheus metrics Built-in exporter, example alert rules Help GET /metrics, tenant-scoped via token Code No native exporter found Vendor
Encryption AES-GCM, customer-held key Trust Envelope encryption, BYOK/KMS, WORM, key rotation Docs AES-256-GCM Vendor
Multi-user, audit None, single binary Global roles, TOTP-MFA for Owner/Admin, audit export owner, admin, operator, viewer, auditor Docs Workspaces, roles per database, audit log, 2FA Vendor
Air-gap egress No Yes Air-gap mode with central egress lock Docs No Vendor

Where Resistro is ahead

  • Point-in-time recovery for MySQL and MariaDB. Databasus offers logical backups only for these engines.
  • Structural restore drill. PostgreSQL: valid GIN indexes, consistent sequences, valid foreign keys. MySQL/MariaDB: indexes, foreign keys and AUTO_INCREMENT counters. Not just row counts.
  • Prometheus built in. Exporter and alert rules ship with the binary.
  • WORM and KMS. Resistro Enterprise encrypts PITR artefacts with BYOK/KMS; on S3 buckets with Object Lock enabled, artefacts additionally get WORM (immutable) retention. Databasus lists AES-256-GCM only.
  • Four-eyes approvals with PDF proof and a tamper-proof audit log.
  • Air-gap egress and table-level PITR. Resistro Enterprise runs air-gapped with a central egress lock and restores individual tables; Databasus lists neither.

Where Databasus is ahead

  • MongoDB support. Databasus also backs up MongoDB (logical only); Resistro supports PostgreSQL, MySQL and MariaDB.
  • Web UI in the free tool. Databasus ships its browser UI under Apache-2.0; the free Resistro CLI has none (Resistro Enterprise is a commercial product).
  • Helm chart. Databasus ships a Helm chart. Resistro CLI ships raw Kubernetes manifests (deploy/kubernetes), Resistro Enterprise ships docker-compose (deploy/compose) — no Helm chart on our side.
  • Multi-user access in the free tool. Databasus lists workspaces with per-database roles, an audit log and two-factor login; the free Resistro CLI is single-user. Resistro Enterprise has global roles and mandatory MFA for Owner/Admin.
  • Price. Databasus is free; Resistro Enterprise is a commercial licence.
  • Release pace. Databasus ships new releases at short intervals.

Measured against the dump-tool baseline

The Databasus table compares feature scope. Raw dump and restore speed we measure against mydumper and myloader instead: on backup we are slower than mydumper (about 1.11× on large databases, 1.6–2.0× on many small tables) and clearly faster than mysqldump; on restore we are on par with myloader on multi-table datasets (0.92–1.04×), 1.13× slower on the 9.7 GB / 29-table measurement, and 1.395× slower on an older single-table run. All figures: docs/POSITIONING.md.

Transparency note

Vendor rows come from Databasus’ public README, retrieved 25 October 2026, and were not run by us. Help, code and docs are weaker evidence than run. Both projects move fast: please check the repositories before deciding. If you find an error on either side, write to us and we will correct the page.