This metaphor about executives seeing all the data but not being able to use it is a very old battle cry in IT. Data comes in faster than ever, has greater density every year and the crisis only deepens. As stated in a previous blog,the network is the nervous system of the enterprise, and as such it must help the stimuli flow. Everyone uses the network to pass information, so what’s the problem? Congestion, priority conflict, bad timing such as sending gigabyte files over the network during regular business hours, and the endless copying of the same data set, so that “I can be in control of my own destiny”.
Sound familiar ? So what can be done about it? How about establishing Information as a Service?
Information as a Service leverages a company’s infrastructure to streamline information flow by consolidating numerous data sources into manageable views and providing that information in real time. Given the typical congestion and chaos, how in your average enterprise network how would such an idea be possible? Remember another previous blog- the 100 million dollar blade? The case was made that change happens no matter what, so the only choice is to think about the business demands and make some intelligent bets while embracing change.
Making endless copies of data at the end of the processing day so that reports can be run is not a long term winning strategy, so it is important to consider a deliberate investment in Information as a Service. To create this service, the designers have to take a Demand Driven point of view, starting with critical business needs:
· What information is needed when?- Explore events and time critical windows
· How dispersed is that data, how many different sources, formats, time zones?
· Understand what decisions are not being made by executives because the information is not perceived to be there
· What kind of volumes must be processed
· How much data must be altered by the executives as opposed to viewed.
Armed with this kind of information, an Information as a Service pilot can be initiated. OK, now what? Get an army of coders?, Draw 1000’s of data diagrams write massive functional specifications that become outdated even as you print them?, Don’t panic
There is a growing list of products that help an architect create a federated data view, where moving the data is neither required nor encouraged. These products can find structured and unstructured data, build the relationships and then let the designer choose meaningful aliases. There are also emerging products that can help map the relationships between interacting servers in a data center, uncovering non-obvious collaborations between systems. This auto-discovery capability, allows architects to see how data moves, helping them make better consolidation choices. Finally there are advances in data caching and disk technology that make it possible to create tiers of query priorities based upon service levels. Coding does not go away, but it is now part of a sensible approach to leading with the business problem, while utilizing tools and products that add value to this complex problem.
This effort is not painless, and it will cost, but it is proactive as it embraces change, and shows how an enterprising technologist can make change an ally, and promote the network as a digital analog of the business
Sheppard D. Narkier
Chief Scientist & Co-Founder
Adaptivity Inc.
