Free dedupe test for good home

Are you interested in purchasing a target deduplication system (NAS or VTL)?  Do you want to perform full testing before you buy, but have been unable to do so due to lack of manpower?  Did you get a quote from a consulting company on what it would cost for them to do the testing, and almost die of shock?  What if I and my cohorts could do the test for free?  Would that be interesting to you?  Click Read More for details.

Many months ago I referred in a blog post to a business idea that I’m working on.  Here’s the short version: I am starting a company that will provide completely unbiased real-life product data to those interested in purchasing or upgrading their IT technology.  Think Consumer Reports — but for the IT industry.  We have figured out a business model that allows us to take NO money from vendors, but still give them enough incentive to cooperate with our process.  Once fully-formed, the company will be unlike any company I’ve ever heard of in this industry.

My first step towards that end is to provide the industry with the first fully-independent head-to-head test of target deduplication products.  I plan to test several of them using real data in a real datacenter and then prepare a fully-detailed report on each of them that will be available for purchase.  Pricing details will be available soon, but suffice it to say that purchasing all of the reports will cost a fraction of what it would cost you to perform your own testing of one product — let alone head-to-head testing.

Didn’t I say something about free?

To make this happen, I need to perform the testing in a real datacenter.  Would you like to be that datacenter?  If your datacenter meets the requirements of the test and you’re willing to allow us to perform the tests onsite, we’ll do that testing for free. In addition, you’ll get input into the test, including which system(s) we test first.  (You get free testing; we get a report we can sell; everybody’s happy.)

The way I like to test target dedupe systems is to copy production backups into them.  We use a spare server as the backup server, and spare tape drives (or an entire tape library if possible) to copy your actual production backup tapes (or copies of them) into the target dedupe system.  This allows us to use actual backup data without putting your actual backup system at risk.  It also allows us test multiple target dedupe systems with exactly the same data.

So here’s the kind of datacenter that would make the ultimate place to do such a test.

  • A full backup of your entire datacenter would be at least 10 TB
  • You perform both full and incremental backups on all your data (Yes, I realize this excludes TSM shops.  Sorry about that.  I want to get you on the next test.)
  • You are backing up both Windows and Unix servers, as well as Exchange and Oracle databases
  • Other databases and applications would be a plus (SQL Server, SharePoint, VMWare ESX)
  • You’ve considered buying a target dedupe system (NAS/VTL) but you haven’t pulled the trigger
  • You have the physical resources to do the testing
    • Spare server to act as backup server/data mover
    • Spare tape drives, tape library, VTL, etc.
  • You don’t have the manpower to do the testing  (That’s kind of the whole point, right?)
  • You will allow us access to your datacenter and data, and allow us to share the results of the test (Your company’s name will in no way be connected with the report.  Only we will know where the test was performed.)

If you’re interested, please send me a private email.

Written by W. Curtis Preston (@wcpreston), four-time O'Reilly author, and host of The Backup Wrap-up podcast. I am now the Technology Evangelist at Sullivan Strickler, which helps companies manage their legacy data

2 comments
  • You’ve mentioned that TSM isn’t an option this first round. To keep people with Amanda, Bacula, tar, and the like from coming forward, what products do you want to work with?

  • The only reason TSM is out is that they don’t do full/incremental backups, and that will skew the dedupe numbers and performance. Since the bulk of the market still does full/incremental backups, I’d want the test to reflect that to give it the most value to the most people.

    I doubt that Amanda/Bacula/tar users are going to have 10+ TB of data, but I could be wrong. I’ll cross that bridge if I come to it.

    If I had my druthers, the prospective shop would be running NetBackup. Going with “most value to the most people,” it’s certainly the product with the most market share, even if it is shrinking these days. But that’s not to say that I won’t do this with a NetWorker, Data Protector, or CommVault shop if that’s who volunteers.