Over the past few blog posts, we have discussed the issues that the TSB Bank in the UK has faced (and still is facing several weeks after the initial failure). As a reminder – TSB were moving from a set of legacy applications to a new set of applications. The move appeared to go well but as soon as they “went live” there was a complete meltdown and customers could not access accounts, incorrect information was sent out, payments were not made etc.
Now, the real surprise about this is the fact that the press seemed to be surprised that a bank was having technical problems. Despite the fact that TSB is by no way the first bank to have problems when migrating from legacy systems and it certainly won’t be the last.
When we look at the application architectures of modern banks we have become used to immediate access to a “friendly” interface on any device of our choice – laptop, tablet, phone. What the consumer doesn’t realize is that what we see are amazing new user interfaces but the underlying systems are still, in most cases, the mainframe applications written decades ago in languages that are old with limited numbers of programmers available.
In 1999 the world was getting ready for the disaster of Y2K and companies were spending huge amounts of money and employing thousands of contractors to try and make sure the “planes wouldn’t fall out of the sky” on 1 January 2000.
Surprisingly nothing much happened.
Now we are facing a real challenge – a huge number of organizations are 1 retirement away from disaster. When the last person who understands the language an application is written in or has that vital piece of knowledge locked in their brain retires what do you do?
Back to the banks – most of the back-office applications were developed 20+ years ago written using languages that are no longer taught. Most run in batch mode overnight. Why does your bank give you until midnight to correct an unintentional overdraft – because that’s when the batch processing starts and they can actually charge interest.
The solution is for organizations to run migration projects to move from the legacy environment to the more modern (trying to avoid the new applications becoming legacy themselves).
But, the problem most IT departments face when looking at this type of project is they don’t have a ground zero. They don’t know what applications they have and even if they did they don’t know what those applications do.
Organizations all over the world will have been reading the TSB saga and heaving a huge sigh of relief that today it isn’t them in the newspapers or on television. But, deep down they know it could well be them tomorrow.
Y2K was a walk in the park compared to the disaster that is waiting to hit a huge number of organizations not least 90% of the financial institutions in the world.