Another industry event; this time its HP Software Universe in Barcelona, and again it’s bigger with more attendees, (around 4000 people), than in previous years. This fits a pattern I have seen over the second half of this year with industry events getting better attended again after a period when interest and attendees had declined. It seems that people are realising that there is a sea change in technologies and their use in business going on so they again want to attend to find out more. The crowded stands at these events are the demonstration booths where one person provides a one on one introduction and tuition on a new technology. As I have said before I don’t intend to cover HP products and propositions as they can do that for themselves, but it’s a case of the discussions and linkages that attendance produces.
There is a common feature that I have seen at these events leading me to my real interest. These popular demonstrations all seem to feature the use of so-called easy to ‘drag and drop’ techniques by which you reconfigure your business process through the orchestration of the ‘services’ in the SOA layer. Looks great, probably is as well, but the consequences are the real issue. The HP message at the event is around how to support this with smart infrastructure, i.e. self configuring automatically to handle the impacts of these changes, as well as unpredictable user demands. The use of more Web 2.0 over SOA solutions, possibly as ‘shadow IT’; meaning that the first that the IT department will know about the deployment, or change to a process, is when a ‘drag and drop’ change actually occurs. Therefore the infrastructure needs to be smart enough to identify this and compensate through automated optimisation reconfigurations. Nice thinking, guys, and good to see some real capabilities to handle this situation. But is it enough?
My challenge is that there is a similar, but different problem in the business process stack and I am not sure how this will get handled. Process integrity leading to data integrity – what I will call ‘procedure’ – has taken a long time to achieve in our deployments, and now the ease of change capability looks as though it may threaten this. I accept it seems likely the various software giants have thought of this, though I have not seen any evidence to support that they have, and made some controls. However these new ‘services’ do not behave like applications, or ERP systems, where integrity means achieving a standardised data model, usually in line with Generally Accepted Accounting Principles, GAAP.
No, these new ‘services’ are there to achieve strongly differentiated abilities for your enterprise to do better business in today’s competitive markets. So it’s a little difficult to come up with an effective kit of built in rules and policies for checking deployments when the whole reason for the change is to be ‘different’ to the rest of the market.
So I am left wondering about this; after some thought my conclusion is that it reflects the difference between the existing ‘transactional’ IT systems, exposed as services, and the concept of ‘Interactions’ representing this new techno business environment. My natural instinct is to think of control, and consequences in terms of ‘transactional’ systems, where as may be at the interaction layer we shouldn’t think this way? May be if we have an active smart infrastructure to keep things operating we don’t have to worry about anything else? After all we don’t seem to need these rules for Web 2.0 social networks and their gadgets and widgets.