The XAM announcement today is a major deal for Centera. It caused me to think about conversations that were occuring outside my cubicle three years ago. I sat near the GM and CTO of Centera. Was I eavesdropping? (Hey, the door was open!). This is what I heard:
“Some customers are hesitant to purchase Centera because there’s no second source.”
Fast-forward less than three years.
EMC just announced an API that allows customers to choose someone else.
If you don’t know what XAM is, here’s a brief evolution (from my “Centera” point of view).
- In 2002 EMC introduced its fixed-content storage device for unchanging digital information: Centera.
- Centera created a new market known as CAS: content addressable storage. There was only one way to talk to Centera: through EMC’s proprietary API (the Centera SDK).
- Hundreds of application partners started coding to this API; it was the only API in the industry that was specifically designed for fixed content.
- XAM was proposed to SNIA in December 2005 as an ‘open” API for fixed content. “Open” can be defined as “supported by multiple vendors”.
- XAM was demonstrated at SNW in October 2007 with multiple vendors participating.
We’ll get to the reasons why EMC would do such a thing (support an “open” API) in a moment. First I’d like to register my reaction to the announcement, which falls into two different categories:
- That was fast.
- The floodgates are open.
Executing at the Speed of a Standards Body?
Read the above timeline one more time. XAM was proposed and demonstrated in less than two years. Think about that. A group of storage vendors (competitors) got together to approve an open API for fixed content. Within two years a XAM plugfest was occuring. XAM objects written to a Centera could be read and re-written to HP storage. XAM content stored on Sun hardware could be moved to Centera. Keep in mind it was just a demonstration, but it showed that applications coding to the XAM API will have a choice of underlying fixed content storage devices.
What’s amazing is the fact that a standards body moved that quickly. Some might be saying “less than two years is quick?”. Well, going from proposal to demonstration in less than two years is the speed of light for some standards efforts. And on the heels of the plugfest, EMC made today’s XAM announcement . Sun also announced recently that they are donating their own code as part of a separate effort.
The Fixed Content Floodgates
Back to the aforementioned hallway conversation – why would EMC give up the so-called vendor lock-in advantage that supposedly came from introducing the most unique, fixed-content product in the industry? Why open up an API so that more vendors can get a piece of the huge fixed content market?
Here’s the way I see it. EMC Sales identified customers who said “I want Centera”, but decided against the purchase because there was no second source for Centera. EMC needed a second source to enable more Centera deals! What were the odds of getting storage vendors to integrate with the Centera SDK? Not good. What was needed was a new API for fixed content that could build on the groundwork laid by Centera.
It’s here, and it’s called XAM. Here are some of the features XAM supports:
- content retention
- immutability
- tamper-proof authenticity
- location independence
- ease-of provisioning and capacity expansion
Let’s say you’re a customer that wants to purchase a fixed-content storage system. Now you have choices. I’d like to ask which product you’d choose?
How about a storage system which created the fixed content CAS market?
How about a storage system with six years experience handling fixed content?
How about a storage system which has implemented immutability, authenticity, and location-independence from day one, and retention since 2003?
How about a storage system that’s hardened? I’m talking about years of fixed defects, performance enhancements, and accelerated failure recovery. It’s not a virgin product by any means.
How about a storage system that has mastered the art of scaling to huge object counts?
That’s why EMC supported XAM. Now that this last roadblock (vendor lock-in concerns) has been removed, customers will now choose fixed content storage systems more often. They now have second sources. Fixed content storage system adoption will rise.
And Centera will rise with it.
Steve
