From One-Size-Fits-All to Custom-Fit for IT Ops
Source – devops.com
Here is how enterprise IT Ops used to work: you would make your choice among the Big Four vendors and buy in (quite literally!) to their strategy. This meant a suite of products that came preintegrated (more or less) out of the box, and—at least on the sales proposal—covered all of an IT organization’s needs across a huge spectrum of activities.
Broadly speaking, each of the Big Four’s suites included some flavor of the following: monitoring, ticketing, configuration management and asset tracking and automation and orchestration. Some vendors were stronger in one area than in another, but unless you had pretty specific needs, each vendor could be your one-stop shop.
Development and improvement inside these suites was mainly incremental and occurred at a fairly leisurely pace. Meanwhile, startups with bright ideas would come along occasionally and challenge the incumbents in specific areas, only to be bought by one of the mega-vendors if they showed enough promise.
There was a symbiotic relationship between vendor and IT buyer, as IT professionals built decades-long careers as experts in this or that vendor’s technology stack, with the wall of framed certifications to prove it. CIOs were reassured by the fact that there was a readily available pool of experts to hire, and continued reinvesting in the next iteration of their chosen vendor’s stack.
The Old Model No Longer Fits Modern IT
In the last decade or so, this comfortable status quo has broken down. The cracks were already showing before then, but only the most agile and forward-thinking organizations were moving fast enough to trip over them. As more and more businesses began to expect greater and greater responsiveness and speed from their IT departments, the huge sprawling suites of the mega-vendors were no longer able to keep up with the required rate of change. Leisurely 18-month release cycles no longer cut it in an agile world, where that time span may encompass several entire strategies.
In the old world, to call something “best of breed” was a sniffy dismissal of a one-trick pony, as opposed to the more supposedly versatile offerings of the Big Four. In the new world, it became obvious that aside from a handful of tentpole products that were genuinely good and fit for purpose, most of the individual components of the big suites was distinctly unexciting. Worse, the main value proposition of the suites—that they came preintegrated out of the box—started to break down as well. Many IT professionals were still fiercely loyal to individual products, but they became more willing to pick and choose, rather than buying into the whole vision.
In parallel, the rise of simple integrations—primarily REST APIs—made it easier than ever before to connect disparate solutions together into a single IT Ops toolchain. Worse (from the point of view of the incumbent vendors), sometimes it was easier to connect modern point solutions than their own component pieces, weighted down by the baggage of decades on the market.
How to Deliver the Flexibility and Agility the Business Expects?
These days, it is becoming commonly accepted even among the most staid and risk-averse organizations that the way to go is no longer to align to a single IT vendor and trust them to deliver on their vision, but rather to design custom-fit, multisolution toolchains, including home-grown, open source and commercial tools.
But what to do if you find yourself dealing with the consequences of this sort of IT monoculture?
Look Back At How You Got Here
The first step is to re-evaluate past buying decisions and strategic alignment. It is very important to do this without assigning blame; it was a different time, and those choices may well have been the right ones for the prevailing environment at the time, when integration between tools from different vendors was legitimately difficult.
Don’t Just Throw Everything Out
In many environments, too, the correct solution may well not be a clean sweep. One or more of those tentpole products that kept the overall suite looking good past its sell-by date may well still be perfectly fit for purpose and not worth the hassle of swapping out. However, other components are probably legitimately lagging behind the state of the art, in which case it is worth replacing them with something more fit for purpose.
Count Costs – All Costs
Include cost in these calculations, but remember to include all costs. Maybe you or a predecessor signed a big enterprise license agreement (ELA), so Product X is “free” in that context. But what does it cost to keep Product X running, in terms of ancillary licenses, hardware requirements and, last but not least, skill and effort on the part of the IT administrators? I don’t want to name specific products or vendors as examples, to avoid the discussion devolving into particulars, but it’s worth bearing in mind that many legacy products have shallow talent pools—not that many people are available on the market, and many of those people are coming to the ends of their careers, which makes them both expensive and reluctant to change employers. Instead, up-and-coming technologies often have much deeper and wider talent pools, with many more experienced specialists available at all salary and skill levels. If you don’t believe me, try hiring half a dozen COBOL programmers and half a dozen Node.js coders, and then we’ll talk …
Your IT Ops Toolchain, For Your IT Ops Needs
Moving from one of the big all-in-one suites to an assemblage of different tools, all flying in loose formation, is not easily done. Any technology transition of this sort is a Project with a capital P, and needs to be planned and managed with due care and attention. However, there is a large payoff at the end in terms of increased agility and flexibility within IT Ops. In turn, this affords IT the ability to offer their business counterparts the sort of responsive and even proactive IT support that moves IT from being a cost center, on a par with Facilities, to a strategic differentiator that enables new business initiatives.
Business as usual doesn’t cut it any more, and assumptions that used to be valid have fallen apart. Decisions based on those assumptions need to be re-evaluated. New, more flexible approaches will set up IT Ops for success in the newly flexible world we are already living in.