As organizations move from experimenting with big data to carrying out big data projects, it’s no surprise that one of the first places they look is the data warehouse. In a big data world that’s full of big promises and big complexity, the warehouse environment provides a tangible, logical, known starting point. You bring in Hadoop, offload the data and voila, you’ve lowered costs and made your warehouse more efficient.
Don’t get me wrong, I like lower costs and increased efficiency just as much as the next gal, but I liken this to those ever-present weight loss commercials: Drop 25 pounds in one week with LoseQuick! It’s essentially a weight loss program for your data warehouse. Move all that complicated, unstructured data over to Hadoop to discover a whole new data warehouse—one that’s quick, trim and cheap. But is that really the ultimate end-game in a big data world? More importantly, does it provide long term competitive advantage? No. Long-term success requires more than just an offload; it requires a true modernization of the data warehouse.
First, let’s put this in perspective. Technology today is allowing us to consume more data and generate more insights—insights that are sparking rapidly evolving demands and analytic requests. Companies want access to insight and they want it immediately, but this presents a real challenge. A recent study by Aberdeen showed that 40 percent of companies interviewed thought that the volume of data is growing too rapidly for their current data infrastructure. This drives the need for data warehouse modernization.
So what is data warehouse modernization? It’s the use of new technologies to increase performance, leverage big data analytics, build on, but not replace, existing investments and deliver insights faster. These technologies can include Hadoop to provide exploratory or ad hoc analysis, a query-able data archive and a lower cost for select workloads. Or they may include in-memory and columnar technology to boost performance of analytics.
A modernized warehouse is able to handle new data sources or additional capacity that can support new lines of business or drive new analytic capabilities. It can accelerate analytic queries to enable more timely business decisions or increase performance. It can handle streaming data sources to provide real-time analytic processing. It can do all of this while ensuring you never lose confidence in the data. And it can do it whether it’s through an appliance, software or cloud services.
It’s more than a quick fix weight loss plan. It’s a foundation for success, for innovation, for sustained competitive advantage. How do you know if it’s right for you? And if you find yourself asking questions like, What do I need to do to my infrastructure to handle big data? How can I deliver faster analytic queries? Should I use columnar or in-memory technology for my analytics? How can I reduce costs or get better price/performance? How can I scale new applications without affecting performance? What about the data I have to dismiss because I can’t process it? then you may be ready to modernize your data warehouse.
For more information
- Bringing Value to the Data Warehouse with Big Data & Analytics
- The Next Big 'H' in Big Data: Hybrid Architectures
- Live from Strata 2014: Anjul Bhambhri speaks on bringing analytics to big data with In-Hadoop analytics