July 13, 2009
In May, I blogged about NetApp's announced acquisition of deduplication pionneer, Data Domain. The announcement triggered an unsolicted counter-offer from EMC, followed by another counter from NetApp. But after a month of offers, counter-offers and regulatory reviews, EMC ultimately outbid NetApp with an all cash offer of $2.1 billion. I believe that Data Domain would have been a better fit in the current NetApp portfolio; it would have been easier for NetApp to reposition its current VTL as a better fit for large enterprises that still planned to leverage tape. It's also said that more than half of Data Domain's current employees are former NetApp employees so there would have been a clear cultural fit as well.
For $2.1 billion, EMC gets Data Domain's more than 3000 customers and 8000 installs but it also gets a product that in my opinion, overlaps with its current Quantum-based disk libraries, the DL1500 and DL3000. In Forrester inquiries and current consulting engagements, Data Domain is regularly up against the EMC DL1500 and DL3000. EMC will need to quickly explain to customers how it plans to position its new Data Domain offerings with its current DL family, both the Quantum- and Falconstor-based DLs as well as its broader data protection portoflio that includes Networker and Avamar – which also offer deduplication.
In addition to the EMC and NetApp stuggle to own Data Domain, we also saw one backup application vendor after another add deduplication capabilities to their software. The online backup vendors came forward to remind everyone that they have been deduplicating data in software all along. As I completed my current state of the union report on the deduplication market, no sooner had I completed one research interview when another backup vendor would announce new dedupe capabilities. I could have delayed the report forever waiting for a lull in the announcements. I officially interviewed 19 hardware and software vendors for the report but was briefed by an additional 3 vendors in the past 2 weeks. I already know there will be additional vendor announcements in the coming months.
It was also clear as I was writing the report that the definition of many deduplication characteristics are blurring such as inline vs. post-processing and source vs. target. I took hard stances on some of these definitions and forced vendor offerings into one category or another for the sake of clarity but I also tried to explain the nuances of some of the vendor offerings. For IT professionals, at the end of the day, you must understand essential deduplication characteristics and considerations and familiarize yourself with all the vendor offerings in the market but at the end of the day, you must have well defined performance, capacity, scalability, manageability and resiliency requirements. You must issue an RFI and/or RFP, let the vendors make the case for their offerings, short list 2-3 vendors and bring them in for a proof of concept (POC) to validate their claims. Refence customers and POCs are the best way to cut through all the hype, claims, and counter-claims made by the vendors in this market.
Check out Stephanie's research
You should follow me on here