“Our DataDomain unit has significantly reduced our backup and restore burden. Tape usage, which previously required weekly manual intervention, is now reduced to an occasional rotation for off-site storage. Even with our wide variety of data, including everything from e-mail to already compressed data sets, we are seeing an aggregate 10x compression ratio. The increased capacity and flexibility our DataDomain unit provides allows us more backup options for our departmental users and has made backups and restores an order of magnitude faster.” — Scott Karlin
Manager, Computing Facilities, Department of Computer Science, Princeton University

Compare Data Domain

Data Domain systems are designed to minimize tape for nearline applications such as backup, archiving and networked DR. Other categories of products are also designed for backup to disk, remote office backup, archiving or remote office WAN acceleration, but they have very different design goals and limitations. It is important to keep these in mind as you consider your options. While conventional VTLs, bandwidth optimizers, remote backup software and CAS systems all do what they are designed to do, they do not minimize tape across the distributed enterprise.

There are two kinds of deduplication: those that companies have put into production, and those that are simply not ready. Only Data Domain delivers a complete solution to tape minimization in the distributed enterprise. The differences are easy to find. More importantly, the results are easy to prove – ask for references from your industry before you make a commitment to any vendor.

Learn more about any of these common approaches to nearline storage:

Backup: Primary
Data at Site		 ApproachGoalsLimitations
> 200 TB Traditional VTL:
cache to Physical Tape Library (PTL)		 Risk management; don’t unhook a big investment in tape infrastructure; per-system scale High cost for retention, and replication; built to augment tape not minimize it

Supports backup applications only, no other nearline uses
< 200 GB Bandwidth Optimization:
WAFS / WDS / WAAS / …		 Move data to central data hub

Slow restore speed over WAN only works for smaller sites

Too expensive for very small sites

< 100 GB PC Replication, Remote/Branch Office (ROBO) backup Move protection data to central hub Rip/replace backup software, must support multiple applications

Does not scale up well

Silos of deduplication
ArchiveContent Addressed Storage (CAS) Disk archiving Too slow for backup / restore

Minimum support for backup applications

Deduplication is a Storage Fundamental

It’s obvious that Data Domain deduplication works – economically and operationally. You can see for yourself by asking a nearby customer. But it only works if it is part of the system’s fundamental design. It’s like color in a TV set – you can’t really add it, you need a different design. It’s the same with deduplication.

Adding dedupe as a secondary design element will reveal itself in elements of manageability, resilience, scalability and cost-efficiency. Test it, on your own, without the vendor’s personnel trying to hide the problems. What you’ll find are the seams; these systems do not provide simple, cost-effective tape minimization.