Storagebod And here we go again, another IT systems failure at RBS.
RBS appears to have been having a remarkable run of high-profile core-system failures, but I suspect that it has been rather unlucky or at least everyone else has been lucky. Ross McEwan, the new chief executive of RBS admitted to us in a canned statement that "decades of under-investment in IT systems" is to blame.
Decades seems to be an awful long time, but may well be accurate; certainly when I started working in IT 25 ago, the rot had already set in. For example, a retail bank where I once worked had its core standing order system written in pounds, shillings and pence with a translation routine sitting on top of it – yet many of these systems were supposed to have been rewritten as part of the millennium-bug investigations.
Never mind... pass me the sticky tape
However, most of this didn’t happen. Whole-scale rewrites of systems that were decades old – when only a few people understood how they worked – were simply not a great investment... It seemed companies were satisfied with just patching it up and moving on.
RBS is not going to be the only large company sitting on a huge liability in the form of legacy applications; pretty much all of the banks do as well as many others in other industries. Applications have been moved from one generation of mainframe to the next and they still generally work, but the people who know how they really work are long gone.
Yet this is no longer constrained to mainframe operations; many of us can point at applications running on kit which is 10 years or more old on long-deprecated operating-systems. Just talk to your friendly DBA about how many applications are still dependent on Oracle 8 and in some cases even earlier databases. Every data centre has an application sitting in the corner doing something no one quite understands... and no one wants to turn it off just in case.
Faced with ever-declining IT budgets – either by way of a real decline or being expected to do more with the same amount – legacy applications are getting left behind. Yes, we come across attempts to encapsulate the application in a VM and run it on the latest hardware, but it still does not fix the legacy issue.
If it ain’t broke, don’t fix it ... but the thing is, most software is broken – it's just that you’ve just not yet come across the condition that breaks it. Now the condition that breaks it may well be the untrained operator who does not know the cunning work-around to keep an application running; work-arounds simply should not become standard operating procedure.
Question is as we chase the new world of dynamic operations with applications churning every day: who is brave enough to argue for budget to go back and fix those things which aren’t broken. Who is going to be brave enough to argue for budget to properly decommission legacy systems, you know those systems who only have one user who happens to have a C at the beginning of their job title?
Now it seems that Ross McEwan may be one who is actually being forced into taking action; is anyone else going take action without a major failure and serious reputational damage? Or are you just feeling lucky, punks? ®