Your database.
Where you need it.
A cloud source. An internal destination. One small PGSync runner between them. Choose a route, download your setup, and check the connection from your own host.
Non-sensitive data previewPGSync is in controlled beta. Use synthetic or approved non-sensitive data only. The kit runs on your host; managed hosting is still in development.
Point it at the right places.
Addresses only. Enter credentials privately in the terminal when you run the kit. Nothing entered here is saved or sent to an API.
Inside your network
The runner connects outward to your source and privately to the destination. You don’t need to expose the destination to the public internet. Your source must allow this runner’s connection.
Review the generated PGSync configuration
Add both addresses and your tables to preview the configuration.
The right tool, at the right moment.
Included in your kit- PgBot
- Check schema diagnostics with a separate observer login. No automatic tuning or stored baseline.
- PgTerm
- Open a terminal workspace using an observer account and a read-only profile.
- PGFake
- Mask a separate developer JSON export with an explicit policy. PGSync does not mask rows in transit.
python3 run.py diagnosepython3 run.py inspectpython3 run.py mask --helpA place to prove
the restore, too.
PGRun could power a paid Recovery Lab: a short-lived native PostgreSQL environment for restoring a retained backup, checking it, and cleaning up. A live replica remains a separate thing.
Planned option · provider estimate only · no hosted project or paid tier activated
Published PGRun pricing500 half-hour branches and a 10 GB Safe Copy: $24.70 before changed-branch storage and other costs.
Native engine availability, backup import, region, TTL, and cleanup must be qualified before launch. PGX is for disposable development workloads; it cannot be the only restore verifier.
Sync keeps a copy current. Retained backups let you recover an earlier one. Explore restore verification