zx Object moves S3-compatible object data at the cloud provider's ingestion rate — roughly 10× the throughput of typical movers like s3cmd and rclone. Multipart transfers at high concurrency saturate your links instead of trickling objects one at a time. Object↔object and file↔object, within and across cloud providers.
zx Object is one face of the unified Zettar zx data mover — the same scale-out engine behind file movement, single-site, and streaming, pointed at S3-compatible object storage. One mover, one operating model, every endpoint you run. It runs at the data rates the cloud provider supports and scales out by adding servers — no software ceiling, filling very large pipes — the cloud provider's ingestion rate is the limit, not the software.
Native S3-compatible multipart movement splits large objects into parts and moves them in parallel — so one big object never caps your throughput. Every transfer is verified with checksums and protected with TLS.
Thousands of objects and parts stay in flight at once to maximize pipe utilization — about 10× faster than s3cmd or rclone.
Move between object stores, or between file systems and object storage, through one engine. No separate tools, no staging detours, no format wrappers.
Run inside one provider or move data across providers. Migrate, replicate, and consolidate object storage on your terms — without lock-in.
"Zettar moved an actual petabyte over a 5,000-mile network loop in 29 hours — with encryption and checksumming — at 96% bandwidth utilization."
The object-movement engine inside the unified Zettar zx data mover — S3-compatible multipart movement for object-to-object and file-to-object transfers, at high concurrency and petabyte scale.
It runs at the limit of the cloud provider's ingestion speed and scales out by adding servers, with no software ceiling. Where your environment is capable, it fills very large pipes — your network is the limit, not the software.
Yes — run inside one provider or move objects across providers and on-premises, in both directions. Migrate, replicate, and consolidate object storage without lock-in.
Native multipart movement and high concurrency keep pipe utilization very high — roughly 10x faster. In a real KEK-to-AWS transfer, zx ran 5.9x faster than aws-cli.
Every transfer is verified with unconditional end-to-end checksums and protected with TLS.
See how zx Object moves data to, from, and between S3-compatible object stores at high concurrency — object↔object and file↔object, within and across providers.