The Rumor Mill At Work

For some time I've been saying that data de-duplication is one of the most important--if not the most important--technologies to hit the data protection market in years. (Of course, without the advent of SATA drives, we'd likely be having a different conversation, or at least the timing of this conversation would be very different ... pushed out, way out. But that's a separate discussion.)

So, is it really any surprise, then, that the press has gone bananas over de-dupe? No, I don't think so. The media loves to "dish"--and the storage media is no exception. But in the land of storage, the opportunities are few and far between. So, what better to "dish" about than who's dumping whose de-dupe technology for whose? Let's face it: It is as close to Brad-Pitt-dumping-Jennifer-Anniston-for-Angelina-Jolie as this industry is ever going to get. Therefore, when I read two recent Press articles this afternoon about EMC's supposed split with FalconStor for Quantum, I could only chuckle.

Whether the rumor is true or not (no comment here), readers need to remember:

• De-dupe is a feature not a product. It is just one feature--albeit an important one--of disk-based platforms. This puts vendors that offer de-dupe in the driver seat right now, but over time, as de-dupe becomes widely available, vendors will have to look to other features to differentiate itself. Now, this could mean offering multiple types of de-dupe or global de-dupe support or something entirely different from de-dupe.
• De-dupe is not the market; disk-based backup (or data protection) is. The disk backup market is still emerging. Just look at our recent Research (Data Protection Research, 2007), the majority of users (53%) are still using a mix of disk and tape, and a non-significant amount are still in all tape environments. The point: There's plenty of opportunity for mulitple vendors.
• Data volumes are growing exponentially. This has implications for primary storage, secondary storage (e.g.,backup), and near-line storage (e.g., archive). De-dupe will play in all three areas.
• One of data de-duplication's hidden jewels is the effect it can have on remote replication--something you will be hearing a lot more about going forward from many vendors.
• Choice is a good thing--not just within a vendor's own portfolio (e.g., multiple types of de-dupe) but also among vendors.
• Ultimately, what will make or break a specific technology is whether it works or doesn't. The next six to 12 months will be huge from this respect.

I have some specific issues with these articles, but I'll leave that discussion to if or when an EMC/Quantum deal is actually announced. Don't want to add to the rumor mill.