When regarded as a profit center, IT receives the human capital, time and financial support required to focus on innovation. Supporting legacy servers is necessary for sustaining business operations, but it doesn't allow CTOs to find ways to improve how their companies function.
How can your IT department institute legacy system management practices that establish itself as a profit center?
Find a legacy system support partner
There are organizations specializing in legacy system maintenance, support and repair. These companies maintain inventories of HP Alpha, SUNSPARC, VAX and other legacy equipment, enabling them to provide spare parts and refurbish on-premises equipment as needed.
When assessing legacy system support partners, ensure they can do more than replace compatible RAM and other components. Look for businesses that not only service on-premises systems but also:
- Support multi-vendor environments.
- Configure replacement servers to mirror your original environment's functions.
- Develop strategies designed to extend the end-of-life of your equipment.
Ultimately, a strong legacy system support partner replaces service from original equipment manufacturers. That way, you won't be forced into upgrading your hardware every four years or so. Most importantly, such partners free up the time and resources in-house IT needs to innovate and experiment with business-changing applications.
Legacy systems reinforce the 'cost center' issue
Despite the fact that IT can enable organizations to develop and launch new services, products and solutions, most executives view such departments as cost centers.
According to CIO.com's 13th annual "State of the CIO" survey, approximately half of the 722 CIOs and IT leaders who participated in the study said their colleagues regard IT as "cost centers or service providers." There are three ways ways mismanaged legacy systems exacerbate this issue:
- The business may rely on IT veterans familiar with maintaining applications running on SUNSPARC, HP 3000, HP Alpha or other legacy server environments. With regard to succession planning, the supply of young sysadmins familiar with such systems is limited. Therefore, talent possessing the requisite knowledge may be reaching retirement age, meaning the company will keep churning personnel.
- Fiscally, maintaining legacy systems can be a burden. The U.S. Government Accountability Office found that federal IT leaders planned on spending 76 percent of their 2015 budgets on upkeeping and running legacy systems. This situation is relatively common in the private sector as well, meaning IT departments lack the financial support to fund innovative efforts.
- As business needs arise, developers make changes to legacy system codes. Decades-old applications contain thousands of adjustments that were either poorly documented or never recorded at all. In addition, the developers who applied such changes may no longer be with the company.
Given the persistence of these issues, IT has the incentive to port legacy applications running on anachronistic hardware to modern infrastructures, but such endeavors often run over budget and fail to meet timelines. What's the cost-effective alternative?
How to cost effectively port legacy systems
For the sake of argument, let's assume you're running an application on legacy SUNSPARC servers. The system itself is about 20 years old. Over the years, developers, many of whom have since left your business, implemented changes to the application.
As spare parts for SPARC hardware become scarce and support for SunOS is difficult to come by, you need to consider porting the application to an environment running on Red Hat Enterprise Linux. To achieve this goal using conventional methods, you would have to hire a team of developers capable of converting the application so it can run on RHEL. Projects such as these are known for exceeding timelines and going over budget.
The alternative, unconventional option is to use a cross-platform hardware virtualization solution that emulates the SPARC environment. With this approach, an engineer would conduct an assessment of your application's infrastructure dependencies, and configure the platform to mirror the current infrastructure.
Cross-platform hardware virtualization solutions run on standard 64-bit x86 computer systems while retaining SPARC-based code, from SunOS to the products associated with the operating system. Although the platform engineer, with assistance from your team, will have to make some minor adjustments to the OS, layered products and possibly the application, the changes won't be nearly as large as they would be in a traditional conversion approach.
How does the emulator run on the x86 architecture? The platform uses a separate RPM installer to run over a standard Linux distro. In this case, the RPM installer would enable the cross-platform hardware virtualization solution to run on RHEL.
Utilizing system emulators can run as low as $18,999, depending on the composition of your legacy architecture. These costs are minimal compared to application conversion endeavors, which can run into the millions. In addition, its emulators are proven technologies that eliminate much of the risk associated with code upgrades.