bucketpilot
For teams whose S3 buckets are the only copy

S3 backup that runs on a schedule and can actually restore

Back up any Amazon S3 bucket to another bucket, account, cloud or BucketPilot Cloud: incremental every run, every version kept, and a restore that puts files back.

The backup form: source bucket, destination bucket, schedule, and the versioning switch

The problem

Amazon promises eleven nines of durability, and it keeps that promise: the bytes you wrote will be there. It promises nothing about the bytes someone deleted, overwrote, encrypted, or emptied with a lifecycle rule that matched more than it meant to. For most teams the bucket *is* the archive, and there is no second copy anywhere.

> Measured from the product, September 2026: 240M+ objects indexed, 50 TB under management, across Amazon S3, Cloudflare R2, Google Cloud Storage and customers' own servers.

What BucketPilot does

  • •A backup job copies a bucket to a destination you choose — another bucket, another AWS account, Cloudflare R2, Google Cloud Storage, or a BucketPilot Cloud bucket we run for you — hourly, daily, weekly or monthly.
  • •Incremental every run. The first run copies everything; every run after copies only what changed, compared object by object on size and checksum. A bucket of tens of millions of objects is diffed in the database, not in memory.
  • •Every version kept. The destination has versioning on, so an overwrite or a delete at the source becomes one more version at the destination, never a loss. Restore any version, of any file, from the file browser.
  • •Backup health on every bucket. The last ten runs on the card, an email the moment a run needs attention, and a Monday digest that says which buckets are covered.
The backup form
The backup form

How to start

  1. 1.Connect the AWS account under Credentials — a one-click, least-privilege role, or a key.
  2. 2.Open Backups & Migrations, press Create, choose the source bucket and a destination. If the destination does not exist yet, create it from inside the form; a mirror destination gets versioning on and locked.
  3. 3.Pick a schedule. The first run starts now; you can watch it or walk away.

Backup jobs are on the Pro plan and above. Every account starts with a 14-day Pro trial, no card needed.

What it is not

It is not a sync that mirrors deletes. Nothing a backup job does removes an object from the destination; the only thing that ever leaves is what a lifecycle rule you wrote yourself expires. And it is not an agent: there is nothing to install on a server for a bucket-to-bucket backup.

Questions

Can the backup go to a different AWS account?
Yes. Connect both accounts as credentials and choose a bucket in each. Cross-account is the common case for a backup that should survive a compromised account.
Does it copy the whole bucket every time?
No. Only the first run does. Every later run compares source and destination and copies what is new or changed; for a large bucket that is minutes, not hours.
What happens if a run fails halfway?
The run recovers on its own: transient errors retry with backoff, a worker that dies mid-run is detected and the run resumed. Objects that failed are listed by name in the run log.
How do I restore?
Open the destination bucket, find the file (search works across every version), and choose the version you want back. For a whole folder, a restore job puts it back where the backup took it from.

Go deeper

Back up my first bucket

Full Pro access for 14 days. No card. Connect an account in seconds.