I recently blogged about changes in the meaning of business intelligence and how we can better use it to support real-time decision making. This raised a lot of great comments. Larry Ellison made his own impact on this topic with a webcast to introduce the second generation Oracle/Sun Exadata 2.0. This was an early indication of what Oracle hopes to achieve with the Sun acquisition. It raised all the usual comments on this unhappy with the acquisition given the resulting monopoly it would give Oracle in databases, due to gaining control of OpenSQLtopic, even including whether the approval for completion of the deal will go through (The EU is apparently ).
What was announced was the first online transaction processing database using Sun FlashFire technology to provide in a single 1u rack tray as a single cohesive unit. It has eighty flash drives and four SAS disk connections – providing a massive (multi terabytes) memory – with a capacity of one million input/output operations per second. This raises a lot of questions as to what exactly this is and how it works. You can get answers at the official Oracle Exadata FAQs site.
But behind all of this is the point that this is the first move in a shift towards using flash memory and solid state drives to fundamentally change the accessibility of stored data. This is an entire data warehouse ‘as an appliance’ with operational speeds for any dimension increased by 50 to 200 percent and a claimed cost reduction (over conventional methods) of 50 percent. This is potentially a game-changing move for anything that uses data and by that I mean high-speed accessing of data.
The normal situation is that an SQL relational database provides rapid transactional processing but for capabilities such as traditional business intelligence reporting, a cube structure is required. If you want to get into this side of matters in detail an interesting article was recently published on database journal about cubes and storage.
Now: back to business intelligence in the front office. This will mean analysing a lot of data, but it will almost certainly be unstructured. Traditionally that’s a ‘no no’. You need to structure the data and design a cube or similar approach to optimise the search. Put another way, you need to decide on the answer you are looking for before you can ask the question!
We are not yet at the point of having true ‘flat’ data, which would involve the capture of all unstructured information about market, enterprise, events – and offer the ability to find unforeseen relationships. But there is a strong move in that direction and we now need to get our thinking caps on, to see how we can use it to make a real difference with internal business intelligence.
As a pure personal issue – though it has got a bigger message – I would like to make mention of the rebranding of one of the oldest computing professional bodies in the world. The British Computer Society, BCS, was founded in 1957 and has always been the professional body for UK computing people. But the world is changing and with it the definition of those ‘people’ and so in recognition of this it becomes; ‘The Chartered Institute for IT – enabling the information society’. Its role is now described as:
‘BCS, The Chartered Institute for IT, promotes wider social and economic progress through the advancement of information technology science and practice’.
A sign of the times and it’s good to see one of the longest established professional bodies in the computing world moving decisively with the times to stay relevant to a new world of users.