FEATURES / BACKUPS

Backups you find out are good before you need them.

A backup nobody has ever restored is a guess. Depfloy schedules dumps of six data engines and your project files to a destination you own, records what each run produced, and restores from the same screen you scheduled it in.

Start free trial Read the backup docs

Scheduled backups is available on the Pro plan and above. Compare plans →

WHAT AND WHERE

Seven sources, six destinations.

Every schedule is one source going to one destination, so a database can go daily to S3 and weekly to an offsite bucket without either schedule knowing about the other.

Sources

The six engines the runner has an adapter for, plus the files your project writes at runtime.

PostgreSQL Service →
MySQL / MariaDB Service →
Redis Service →
Valkey Service →
ClickHouse Service →
Meilisearch Service →
Project files
Destinations

The dump is streamed from your server straight to the destination. Depfloy never stores a copy.

Amazon S3
Bucket, region and path prefix
Cloudflare R2
S3-compatible, no egress fee from Cloudflare
Any S3-compatible endpoint
MinIO, Backblaze B2, Wasabi and the rest
SFTP
Onto a machine you already trust
FTP
For the older host that is still in the picture
The server’s own disk
Fast, and worth pairing with an offsite copy
EVERY RUN LEAVES EVIDENCE

A schedule that shows its work.

Size and checksum

Each run records what it produced and a SHA-256 of it. A dump that suddenly halves in size is visible in the list rather than discovered during an incident.

Retention that prunes

Keep the last fourteen daily runs and eight weekly ones. Older copies are removed after a successful run, so the bucket does not grow without anyone deciding it should.

Exclusions

Project file backups skip what you tell them to — node_modules, compiled caches, logs. Usually the difference between a backup that completes and one that times out.

RESTORE

The half everyone skips.

Restore targets a database you choose, which means the same feature that protects production is the one that refreshes staging from it. Teams that restore routinely are the ones whose backups work.

acme_production — daily · Acme S3 · keep 14
completed 25.07 13:57 4.5 GB Restore →
completed 24.07 13:57 4.5 GB Restore →
completed 23.07 13:57 4.5 GB Restore →
FAQ

Backup questions

Can I restore, or only download? +
Restore is part of the product. Pick a completed run and the database it should go into — that is what makes refreshing a staging environment from a production dump a routine operation rather than a manual one.
Where are backups stored? +
Wherever you point them: your S3 bucket, Cloudflare R2, any S3-compatible endpoint, an SFTP or FTP host, or the server’s own disk. Depfloy never holds a copy — the runner streams the dump from your server to your destination.
What happens to old runs? +
Retention prunes them. Set how many copies to keep per schedule and the runner deletes the ones that fall outside the window after a successful run.
Can I exclude files from a project backup? +
Yes. Exclusion rules keep node_modules, compiled caches and log files out, which is usually the difference between a backup that finishes and one that times out.
Is MongoDB included? +
No. Depfloy does not provision MongoDB at all today, so there is nothing for the runner to back up. The adapters that exist are PostgreSQL, MySQL, Redis, Valkey, ClickHouse, Meilisearch and project files.

Schedule the first one today.

7 days free, no credit card. Point a schedule at a bucket you already have.

Start free trial See the data layer