What would you use to back up R2 to another S3-compatible provider?

Rclone is a straightforward option for bucket-to-bucket transfers. Wasabi is one destination to consider if you want the copy outside Cloudflare. Run the job on a schedule and make sure failures generate an alert rather than just disappearing into a log.

Be careful with sync settings. A mirror that propagates deletions can be useful for disaster recovery, but it is not the same thing as retaining older backup copies.

First decide whether you need a current replica or actual backup history. If someone deletes or overwrites an object, a current replica alone may not give you an older version to recover. Veeam Backup & Replication is worth evaluating if you need multiple days of retention.

Check the product's support for your source and destination before committing, and test restoring both individual objects and a larger set. The recovery process matters as much as getting the initial copy across.

CloudSoda is another option if you would rather use a tool with automated jobs and email status alerts than maintain transfer scripts. It supports transfers involving S3 and blob storage. I'd check its R2 compatibility and pricing against the amount of data and transfer frequency you need.

AWS DataSync is worth investigating, as are native replication features where they apply. Don't assume an AWS bucket replication setup will work directly with R2 though. Check the supported endpoints and how the copy would be triggered.

Also look beyond the initial setup. You need visibility into missed or failed transfers, plus a way to check that the destination has what you expect.

Storage tier matters here. A separate S3 bucket with an archival tier such as Glacier might suit a long-term copy, but it is a poor fit if fast access is the main requirement. Compare retrieval time as well as storage cost.