For a software engineer, there's really nothing more satisfying than completing a product release and watching it enter the market and do well. An even better feeling occurs when the product release defies expectations and takes off beyond any "best case" scenario.
I've always thought that the growth and success of Centera was due in large part to having the right features at the right time. Events in the marketplace occurred that created a pent-up demand for CAS features. Application vendors integrated with the Centera API and created software that took advantage of those features. And finally, customers began buying hundreds, and then thousands, of Centera systems.
It was like watching a Perfect Storm.
Well, it's happening again.
The First Storm: Retention and Compliance
Earlier this decade, businesses were considering the ramifications of corporate accountability for electronically stored records. This period of time is often called "post-Enron". Governments began passing laws mandating the retention of certain types of "electronic documents". This type of legislation came with hard deadlines for complying with the new regulations. Audits were proposed. Any businesses that were not in compliance with these new regulations could suffer stiff fines. The Sarbanes-Oxley Act is one example of legislation that occured in the United States; but it was occuring world-wide as well.
Meanwhile back at the ranch, it just so happened that the Centera organization was putting the finishing touches on a software release called "Compliance Edition". The Centera API was enhanced to support the association of retention periods with content, and the Centera software itself was enhanced to reject the deletion of unexpired content.
At the same time, application vendors were enhancing their own applications to support the new regulations. For example, email archive vendors were offering the ability to retain emails on Centera. Mortgage software assigned retention periods to loan agreements. Financial software allowed for the retention of quarterly financial results.
Government regulations. New Centera features. Application enhancements. Financial penalties for businesses.
Perfect Storm.
Is it any wonder that in 2003 there were press releases about Centera being one of the "fastest growing product lines in EMC's history"?
The Second Storm: The Digital Universe
This week has seen two announcements that make up two parts of a three-part storm system that is coming together at just the right time.
Firstly, a study was released which predicted yet more explosive growth for fixed content (Centera's sweet spot). The report is interesting, and it is challenging. It is interesting from the standpoint of the enormous amount of data represented by not only personal information but "digital shadows" as well. It is challenging from the standpoint of having to eventually manage quadrillions of objects in a secure and performant way. Many of these objects will be relatively small in size.
Meanwhile back at the ranch (are there ranches in Belgium?), the second part of the storm has been brewing. The Centera team has been busy. Last month I wrote a post about the continual progress that Centera has been making on not only storing huge numbers of objects but rebuilding them quickly in the face of a disk failure. Today marks the second announcement of the week: CentraStar 4.0 has been released. It can store an enormous number of small files (up to 25 million per disk) and rebuild content on failed drives faster than ever before.
And finally, the last part of the storm is XAM. During the first storm, some customers were hesitant to purchase Centera. This was due to the fact that the Centera API was "proprietary". Customers understood that it had to be proprietary because it was brand new technology. But it gave them pause because they had no second source. So in a sense, the first "perfect storm" could have actually been bigger than it was.
EMC recognized this and to its credit pushed hard for an industry standard API for storage systems like Centera. Last October SNIA demonstrated a variety of storage vendors interoperating via the XAM API. So the "proprietary" label is a roadblock that can be removed from the discussion. Customers now have tangible evidence that application vendors can program to the XAM API. XAM means choice and interoperability when it comes selecting back-end object storage systems.
Digital Universe. New Centera capabilities. XAM. Millions, billions, .... quadrillions.
It's starting to get windy.
Steve
Comments