If your organization wants to fully embrace a DevOps culture that’s focused on continuous delivery, you need to deal with your legacy IT infrastructure issues. While having a legacy can often be a boon to a well-established company’s brand, legacy IT infrastructure still faces issues due to software and systems architectures. In some instances, such architectures have been in place for years—possibly decades—and can be a death knell in the age of digital transformation and customer-centric commerce.
Legacy IT in a DevOps culture
Legacy IT is more common in some sectors than others, particularly banking, insurance, and government agencies. All of these are challenged to be more flexible and nimble in the delivery of products and services. Fintech upstarts, for example, are giving traditional financial institutions a run for their money, in part because they’re embracing a customer-centric, iterative approach. More ominous for organizations reliant on legacy IT is the increased likelihood of major disruptions to the business.
As customer expectations become more demanding and more business interactions become digital, many enterprises are finding their systems simply can’t keep pace. Early this year, CloudCraze released its 2017 B2B Digital Commerce Imperative Report, which found that legacy commerce systems are proving to be a barrier to digital transformation for many B2B brands. Legacy IT infrastructure issues are thwarting their ability to deliver flexible and agile buying experiences, and it’s leading to lost business.
Just as significantly, legacy IT infrastructure issues are causing organizations to accumulate “technical debt” as they take a bimodal approach in their digital transformation efforts. As defined by research firm IDC, technical debt is “work left to do.” An extreme example would be an operational meltdown.
Otherwise avoidable disasters are caused by aging systems that have become brittle as a result of budget pressures and deferred investments, according to Mike Rosen, research vice president for executive programs at IDC. The challenge, he says, is that technical debt is often not immediately visible. IT departments struggle to identify technical debt, let alone explain its liabilities and consequences to business executives or a company’s board of directors.
IDC group vice president and IT executive advisor Joe Pucciarelli says bimodal IT compounds technical debt further as enterprises are embracing DevOps for new development but keep it from separate traditional IT at-scale processes. This hampers innovation. Splitting those who support traditional IT from those responsible for innovation required by business demands is creating a divide that has consequences for the organization.
Merging new and traditional legacy IT
Rosen says IT is being split between new and traditional legacy IT—the fast and the slow—and this division creates a chasm. The fast, iterative IT that comes from embracing an iterative mindset and a DevOps approach creates a chasm. Long-term issues that haven’t been thought through by fast IT as it races to innovate land in the laps of operational IT.
Just like financial debt, technical debt must eventually be paid. If it continues to go unpaid it can manifest itself as outages that affect business operations. Many organizations, rather than addressing the technical debt, make changes and enhancements that lead to fragility, inefficiencies, complexities and workarounds. It also creates indirect costs, as well reduces and even eliminates the return-on-investment of technologies that have been deployed.
It’s not that DevOps negates everything that’s been done before. Existing frameworks, such as IT service management (ITSM), remain relevant, and can even contribute to an organization’s agility. Release management, for example, can be a collaborative process between developers and operations in an agile, ITSM DevOps world.
Ultimately, if your organization is embracing digital transformation, you should be leaving legacy IT systems behind, and your DevOps team should be focused on supporting a cohesive cloud strategy that includes onboarding traditional IT systems, as well as shadow IT that your organization has become reliant on. But legacy systems don’t have to disappear; DevOps is destined to disappear too, according to Pucciarelli, as it will cease to become a discrete business unit and be woven into the fabric of the organization as their digital transformation efforts mature.