Are You Talking to Me?
If business services are the stars of modern IT, then legacy systems are a lot like black holes. They’ve been around longer than everything else, and what we know about them is mostly conjecture. Yet despite their inherent uncertainty, they represent the promise of technology: they’ve done what they were supposed to do, and they did it without going supernova.
A recent article in the New York Times Technology section made me a little nostalgic for my own bits of legacy, a long time ago in a building kind of far away, and I’ve been thinking about it a lot since then.
Ten years ago, I was an analyst at a large New York bank. Officially I was the Change Management Administrator for the Y2K project. This meant that every production IT change for any system – mainframes to little NT boxes – went through me. While this seems like an incredible responsibility for such a young man – I was 27 at the time – the position could largely be regarded as a sinecure. This is because process and respect for application lifecycle was deeply ingrained into the corporate culture, and nobody would dare to make even the slightest change without filling out the proper forms, running the appropriate tests, and updating the associated documentation – even with 1/1/00 right around the corner.
In the end, the project went smoothly. The financial world didn’t collapse when the digits rolled over, and we all still had jobs. But we were organized and confident, with time, vision, and clarity of purpose on our side.
Fast forward to now: I’m a consultant at a New York investment services firm, and legacy systems are a way of life. We’re engaged in a data-center move, and time is not on our side. “Applications” often consist of Excel spreadsheets talking to Access97 databases, running on single-disk, single-power-supply workstations. Who is using them? What do they do? How much of our business is dependent upon them? And what do we do if we’re forced to disturb them? Anyone who really knew what was going on beyond the Schwarzschild radius is long gone.
But I have an ace in my pocket, and it’s called Tideway Foundation. Using only the operating systems’ own tools, Foundation lets us see past the event horizon and into the core. With a single click, I can generate a report for any server that tells me exactly what processes are running, what version of all related software they are using, what network communications are associated with those processes, and what other machines – workstations or servers – that are on the other end of those network communications. But it doesn’t just stop there. I can even see the processes and bits of software on those remote hosts that are the endpoints of those communications. In short, I now have complete end-to-end visibility of all this system’s functions. I know who’s using it, how it’s being used, and how often. I can print this report out and pass it around to my application owners, and they can then either reconcile this information with what they thought they knew, or incorporate it into a new perspective.
Intelligence like this is a real coup for my customer, who now has new confidence in their ability to deal with systems that were previously thought to be hopelessly obscure. Furthermore with new understanding of these systems, new assessments regarding their criticality and upgradeability can be made at leisure, with great accuracy and little investment in time or money. While legacy systems are our responsibility as IT leaders, their replacement will always ultimately be a business decision, not a technology one – and we need to be ready.
Ten years ago, we could afford to take our time and be methodical. We had all the money and people we needed to get the project done properly. But what are you going to do when gravity starts to pull you, and you can’t outrun it anymore?
RSS