PostgreSQL & MongoDB

Backups you can prove will restore

Most tools verify the dump ran. Norcube verifies it restores.

Automated tests restore your backups into a throwaway database and confirm the data came back. A restore-health score tracks it, and alerts fire when backups fail or stop.

Automated restore testingRestore-health scoreFailure alerts on every planPostgreSQL & MongoDBGzip downloadFree plan, no card

Why Norcube Backup

A cron job tells you the dump ran. Norcube proves your data comes back.

Tested

We restore your backups into a throwaway database we run, check that the data came back, then delete it. After every backup or on a schedule.

Learn more

Anywhere

Managed or self-hosted, on any cloud or your own servers. If Norcube can reach it with a PostgreSQL or MongoDB connection string, it is covered.

Learn more

Alerted

Failed backups and failed restore tests email your team from day one, with Slack or a signed webhook too. Missed schedules get flagged.

Learn more
Restore testing

We test your restores, not just your dumps

Norcube restores your backups into a disposable database, confirms the data came back and tears it down. Run it on a schedule, on a sample of backups or on demand. Setup takes two minutes: paste a connection string, pick a schedule and retention, done.

  • Restore drills into a throwaway database, never the one being backed up
  • Passed badge per backup, restore-health score per database
  • Alerts when a backup fails, a test can't confirm your data, or a schedule silently stops
Try Backup free

Security isn't an add-on. It's the foundation

Your database credentials and backup data are the most sensitive things you trust us with, so they're encrypted at every stage and isolated in every process.

Credentials never in plain text

Encrypted on input, stored encrypted at rest, decrypted only on an isolated server during the backup job. Never transferred unencrypted.

Encrypted archive storage

Archives are stored in AWS Frankfurt with server-side encryption (SSE-KMS), or in your own S3-compatible bucket. Downloads go through signed links.

Isolated execution

Every job runs on its own sandboxed server, with no shared processes and no shared memory. It's destroyed when the job completes.

Made for the day you need to restore

Standard archives you can take anywhere, alerts when something stops, and backups you can start from the CLI or your AI assistant.

From your AI assistant

Ask Claude, Cursor or ChatGPT to back up before a risky migration.

From your AI assistant

Ask Claude, Cursor or ChatGPT to back up before a risky migration.

Over the Norcube MCP server, your assistant can list your databases, start a backup, wait for it to finish and run a restore test on it, with your permissions and no access to the data itself. See Backup over MCP.

No lock-in

Native pg_dump and mongodump archives, gzipped. Restore with or without Norcube.

No lock-in

Native pg_dump and mongodump archives, gzipped. Restore with or without Norcube.

Each backup is exactly what the official tool produces: a pg_dump custom-format file for pg_restore, or a mongodump archive for mongorestore, gzip-compressed. Download any snapshot from the dashboard and restore it with standard tools, with or without Norcube.

Alerts on silence

Flags a schedule that quietly stops, not just failures. Email, Slack or a webhook.

Alerts on silence

Flags a schedule that quietly stops, not just failures. Email, Slack or a webhook.

Alerting works with zero configuration: failures email every verified member of your organization from day one. A dead-man’s switch watches for the more dangerous case, a schedule that quietly stops running, and alerts on absence, not just errors. Alerts are deduplicated (first failure immediately, then a consolidated follow-up instead of an email per run) and a single Recovered notice closes the loop when backups succeed again. Add Slack or an HMAC-signed webhook endpoint, and verify the whole pipeline with one test button.

Storage you choose

Managed storage in Frankfurt, or your own S3, R2, B2, Wasabi or MinIO bucket.

Storage you choose

Managed storage in Frankfurt, or your own S3, R2, B2, Wasabi or MinIO bucket.

Norcube-managed storage lives in EU Central 1 (Frankfurt). To keep archives in your own account, add an Amazon S3, Cloudflare R2, Backblaze B2, Wasabi, MinIO or other S3-compatible bucket; Norcube writes, checks and deletes a test object to verify it before saving.

Any schedule

Hourly, daily or any cron expression, per database, in your timezone.

Any schedule

Hourly, daily or any cron expression, per database, in your timezone.

Production hourly, staging daily, client databases on their own rhythm. Define a schedule and retention once as a policy and attach it to any number of databases; each database keeps its own backup history. Schedules follow your timezone, daylight saving included.

PostgreSQL & MongoDB

Full pg_dump and mongodump backups, MongoDB replica sets and clusters included.

PostgreSQL & MongoDB

Full pg_dump and mongodump backups, MongoDB replica sets and clusters included.

Backup connects with a standard connection string and runs a full logical dump: pg_dump for PostgreSQL, mongodump for MongoDB (standalone servers, replica sets and clusters). Each dump is gzip-compressed and stored automatically, with the same scheduling, downloads, alerts and restore tests for both engines. MySQL and Redis support is planned, with the same workflow: connect, set a schedule, download when needed.

Works with your provider

Managed PostgreSQL and MongoDB connect like any other database. Each guide covers the platform's connection string, network access and managed extensions.

Any other PostgreSQL or MongoDB works too

See all provider guides(opens in a new tab)

Pick a plan. Upgrade when you grow

Every plan has restore testing, the restore-health score and failure alerts. Plans differ in how many databases you protect and how much runtime and storage are included.

  • Free

    $0

    For one database and a first look.

    • 1 database
    • 2 GB of storage included
    • 2 hours of job runtime included
    Start free
  • Starter

    $12/ month

    For a side project or a small team.

    • Up to 5 databases
    • 25 GB of storage included
    • 10 hours of job runtime included
    Choose Starter
  • Pro

    Most popular

    $39/ month

    For production apps with several databases.

    • Up to 20 databases
    • 100 GB of storage included
    • 40 hours of job runtime included
    Choose Pro
  • Business

    $89/ month

    For teams running many databases.

    • Up to 50 databases
    • 250 GB of storage included
    • 120 hours of job runtime included
    Choose Business

Runtime and storage beyond a plan are billed per second and per GB. See the rates and estimate your monthly cost →

Questions, answered

Quick answers to what teams ask us most about Backup. Need something else? Email [email protected].

Which databases does Backup support?
PostgreSQL and MongoDB. Connect any standard instance, whether managed (RDS, Atlas), self-hosted or running in your own cluster, and Backup handles the rest.
How do I schedule backups?
Cron expressions, with timezone awareness. Backup policies bundle the schedule with compression, encryption, retention, parallelism and bandwidth throttle. Attach a policy to a data source and it runs automatically. You can also trigger a backup manually any time.
What backup methods are available?
Logical dumps (pg_dump for Postgres, mongodump for MongoDB): portable, compressed archives you can restore anywhere, in-product or with the standard tools. Physical backups and point-in-time recovery (continuous WAL/oplog archiving) are on the roadmap.
How long are backups kept?
Two retention dials you set per backup policy, enforced together: age (how many days to keep a snapshot) and count (a minimum number of recent snapshots to always keep). A snapshot is deleted automatically once it is both older than the age limit and outside the most recent count. You choose both, so the same data source can have a daily/7-day policy and a weekly/1-year policy side by side.
Can I restore from within Norcube, or do I have to download the dump?
Both. Every snapshot is downloadable as a gzip archive for restore on your own machines. In-product restore is also built in, with three target modes: restore into a new database, overwrite the existing one, or restore onto a new server you provision, each with a pre-flight check that the target is writable.
How do I know a backup will actually restore?
Restore testing proves it. Norcube automatically restores a backup into a throwaway database (one it spins up for the test, with nothing for you to provide, or a scratch server you nominate), confirms the data came back and tears it down. Run it on a schedule, on a random sample of backups, or on demand; a Passed badge and a restore-health score show recoverability at a glance. It's also honest: if a backup restores but one object can't be rebuilt (say, an index too large for the engine), it's flagged as passed with warnings rather than a bare pass, so you see the caveat instead of just a green check. Restore tests only ever restore into a fresh throwaway database, never the one being backed up.
How are credentials and backup files protected?
Database credentials are stored encrypted using AWS KMS envelope encryption. Managed backups are encrypted at rest in S3 with AWS KMS (SSE-KMS); in your own bucket, the bucket's encryption settings apply. Client-side encryption is on the roadmap. Connections to your database use TLS whenever your database offers it. Managed backups are stored in AWS Frankfurt (eu-central-1), or in your own bucket if you connect one.

Protect your first database in two minutes

One form covers the connection, schedule, retention and restore testing. Norcube takes care of encryption, the schedule and the moment you regret deleting the wrong row.

About 2 min
  1. Connect your database

    Choose Protect a database, pick PostgreSQL or MongoDB and paste the connection string. Norcube verifies the connection before you continue, and credentials are encrypted with AWS KMS the moment you save them.

  2. Pick a schedule

    Daily at 2 a.m. in your timezone is the default. Choose weekly, or switch to Advanced for any cron expression.

  3. Choose retention and restore testing

    Pick how long to keep snapshots (Standard keeps 30 days and always the last 14) and how often to test that they restore: weekly by default, daily, or off. Then save, and the schedule starts.

  4. Run it now to checkOptional

    Hit Run now to start a backup immediately and watch the job go from queued to running to success, with a downloadable archive at the end.

  5. Fine-tune with a backup policyOptional

    For more control, a backup policy sets compression (gzip or zstd), the encryption mode, retention, parallelism and bandwidth throttle. One policy can serve many databases.

Stop worrying about database backups

Start on the Free plan. No credit card required. Protect your first database in two minutes.

  • Free plan
  • No credit card
  • Standard formats
  • Download any time
// tick. tick. tick.
0 ticks since founding
Coming soon

Join the PromptHub waitlist

PromptHub is in active development. Leave your email and we'll invite you as soon as it opens, with a free tier and no credit card.

One email when access opens. No newsletter, no spam.