Anyone who has travelled in London during the summer will have cursed the creaking transport system we call the Tube. Why is it so hot and crowded? Why can’t you get a phone signal? How about disabled access? Isn’t the gap between the train and the platform dangerous?
Parts of London’s Tube system date back over 150 years. Carrying 9.5 million people in the first year, the Tube network today serves over 1,107 million passengers.
Commuters compare the Tube unfavourably to the mass transit system in Hong Kong or Tokyo. But those who run London Underground, quite reasonably, point out the difficulties of continually patching, extending and adapting an old system dealing with a load it was never designed for.
Yet as for the idea of closing down the Tube completely and fitting an entire new system – the entire city would grind to a halt! The cost would be unthinkable!
It reminds me of IT systems
Legacy systems on which the company depends are also creaking at the seams, costing more each year to maintain, limited in capacity and always carrying the risk of a shut-down. The patches and adaptations not only get harder to do but become a competitive liability as other companies introduce new functionality and efficiency. Yet when the senior managers think about the risk and cost of replacing the legacy system wholesale, they are horrified. And rightly so. There are numerous examples of enterprise replacement systems which have cost business dearly in performance, money, time and reputation.
Is there any other option?
We believe there is.
Software, thankfully, is not the Tube system. We don’t have to close down the road and railway system while we dig a new tunnel! Instead we can keep the old system running in its entirety while we isolate a single strand of functionality – a single tube-line, if you like – and replace just this.
The clever bit (and where the metaphor collapses) is to ensure both lines work so that the switch over from old to new appears seamless to users.
Sounds easy, right? The problem, of course, is that the few intersecting lines of the Tube are simple compared to the tangled mess which most legacy IT systems represent. It can be a tangle where even the staff charged with maintaining it barely know why or how it works and are terrified to change anything in case it breaks. It’s not unusual to have hundreds of thousands of lines of code with no test coverage, no version control and no documentation.
Specialists in modernising legacy
At Energized Work, we’ve become expert in replacing legacy systems: knowing how to map the mess, create a safety net of tools that make changes safe, disentangle individual strands, extract and replace them one by one.
Eventually the whole big ball of mud is replaced with new services that are understood, run more efficiently and allow companies to take advantages of new opportunities or deal with competitive threats.
It’s much safer, and in the long run, it tends to be much cheaper than any big-bang replacement.
Making it really work
We pioneered and then began specialising in this kind of step-wise approach to modernising legacy systems. We took on a particularly challenging project with Travis Perkins and their Universe based system, parts of which dated back to the late 1980s. The small team achieved an astonishing amount of progress in a short time and demonstrated that a stepwise approach to legacy replacement is indeed feasible. You can read more in the Travis Perkins case study.
That meant we also developed a plan for how to proceed with any legacy system and a set of custom tools for mapping, reporting and testing in particular domains and technology stacks. We have put the detail together into a new White Paper called Solving Legacy.
With 62% of companies believing that maintaining legacy systems offer a negative ROI, it’s time to look at Energized Work’s stepwise approach to modernising legacy.
It’s just a shame we can’t do it with the Tube.
The post Replacing Legacy Systems? Mind the gap! appeared first on Energized Work.