Solutions · Migration

A petabyte a day, not weeks.

Site-to-site, NAS refresh, parallel-file-system, storage consolidation, cloud-to-cloud, and cloud data migration — large-scale data transfer at ~10× typical movers, verified to convergence, with no vendor lock-in. One customer cut migration time by 80% (case study).

The problem

At petascale, the migration is the project

  • Site-to-site and cloud-to-cloud moves drag on for weeks. Single-stream tools like rsync and scp cap out far below your link, so a planned migration becomes a quarter-long project that ties up your engineers.
  • NAS tech refreshes and parallel-file-system migrations risk long downtime. Cutover windows stretch, workloads stall, and the maintenance window never seems big enough.
  • Proprietary transfer and tiering formats trap your data. Once it lands in a closed, proprietary wrapper, leaving means paying to migrate again — vendor lock-in by design.
  • Verifying a petabyte is slow and error-prone. Without inline checksumming you cut over on faith, not proof that the target matches the source.
How Zettar solves it

Move it once. Move it fast. Move it verified.

~10× faster

Line-rate throughput

zx runs at line rate and scales out with no software ceiling, about 10× faster than typical commercial or freeware movers — so weeks of transfer collapse into days.

Verified cutover

Incremental sync to convergence

Repeated incremental passes close the gap until source and target match. You cut over only on a verified convergence — no extended downtime, no guesswork.

Any system

File↔object, NAS, PFS, cloud

One engine migrates file↔file, file↔object, and object↔object — across NAS, parallel file systems, and clouds, including cloud-to-cloud. Consolidate or refresh storage without a forklift.

No lock-in

Native format, exit anytime

No proprietary tiering or transfer wrappers. Your data stays in its native format on open systems, so you are never paying to migrate out of someone else's format.

Proof, not promises
Zettar moved an actual petabyte over a 5,000-mile network loop in 29 hours — encrypted and checksummed — at 96% bandwidth utilization.
— SLAC National Accelerator Laboratory & ESnet, U.S. DOE · Winner, Supercomputing Asia 2019 Data Mover Challenge
80%
less migration time in one large-scale file-migration deployment

Planning an imaging-archive move? PACS and VNA migration at scale →

Build your business case

What is a slow migration costing us?

FAQ

Common questions

What is data migration?

Moving data to new storage, a new system, or the cloud — as a one-time or phased project. At petabyte scale it is the difference between a weekend and a quarter.

How long does a petabyte migration take with Zettar?

About a day per petabyte on a 100 Gbps link, and less as you scale out onto bigger links. zx moves data at line rate — using the bandwidth you already pay for — so migrations stop dragging on.

Will it keep up at petabyte scale?

Yes — zx is symmetric and scale-out: every node both sends and receives, and aggregate throughput grows as you add servers, with no software ceiling. Add servers to fit the maintenance window; your environment, not the data mover, is the limit.

Can Zettar migrate between file and object storage?

Yes — file-to-file, object-to-object, and file-to-object in native formats, with no lock-in.

Will Zettar lock us in?

No. Data moves in native file and object formats with no proprietary wrappers, and the appliance runs on HPE ProLiant enterprise servers.

Does it migrate to and across clouds?

Yes — on-premises to cloud and cloud to cloud, bypassing the 30–50% cloud-stack penalty.

Is data verified during migration?

Every transfer is encrypted and protected by unconditional end-to-end checksums, so any corruption is detected, never silently accepted.

Get started

Move it once. Move it fast.

See how Zettar migrates petascale data between sites, systems, and clouds — verified to convergence, with no vendor lock-in.