As Big Data initiatives mature into enterprise data sources supported by NoSQL products for analytics and operational systems, a clash of cultures is on the horizon (if not here already). Traditional IT implementations teams and their top-down programs rarely see eye to eye with the grass roots culture of NoSQL platform operators. But this divide is not merely between the camps of Big Data/NoSQL and traditional IT implementation teams. This is just the tip of the iceberg…. The divide becomes much more pronounced when you take the discussion to the executive suite. CMOs and CFOs, who “own” results of analytical and operational systems, are less concerned with data center standards and development methodologies as they time to value. CIOs and CTOs, responsible for implementing the connectivity and integration between NoSQL platforms and the rest of the traditional IT environment, are facing pressures to avoid chasing the latest technology fad(s).
Informatica Goal: Maximize Return on Data
The theme of last week’s Informatica Analyst Conference was utilizing the “secular megatrends” of information technology to energize data integration across organizations at an enterprise scale. These megatrends, described as trends we can all agree upon, are the following:
If you look at the history of Big Data requirements (volume, velocity and variety), and the NoSQL platforms supporting those requirements, you see a history of organizations and development teams breaking the mold of traditional information technology (IT) programs. Instead of following the traditional IT methodologies to solve the Big Data issues, these teams pushed the envelope and invented new technologies to solve those “volume, velocity and variety” problems. More often than not, these efforts were accomplished using collaborative, bottom-up methodologies, such as Open Source, rather than rigid, top-down approaches found in traditional product development methodologies. Specifically, if you look at the history of the Hadoop development at Yahoo, you see an approach that sought the input and wide spread resources of the Open Source movement rather than a more rigid proprietary approach.
Many times when discussing the topic of big-data, the focus is on the volume of the data, the structure of the data or the near real-time analysis requirements of the data. We toss around buzzwords like Hadoop, structured vs unstructured, etc…. However, often times what is missed are the analysis goals of the big-data environment.