How automation can reduce technical debt for state government legacy systems

IT leaders will face increasing technical debt if they continue to patch legacy systems and fail to take steps to address underlying IT bottlenecks, according to an IT expert in a new podcast. Automation and orchestration can help, though, if it is applied thoughtfully.

Chris Martin, technology solution architect at Intel, says that by first understanding dependencies in a legacy system, agencies can avoid problems later when introducing automation and taking full advantage of its benefits.

Automation, a tool used to streamline a single process or task, can be a great help to agencies looking for ways to improve their workflow. It reduces the risk of errors and can enable IT to deliver faster updates, patches and code deployments. As a capability, it not only addresses security issues, but it brings cost-savings to agencies as well.

“The benefit of automating processes comes down to the human element, especially in regard to big data and other types of identity access management systems,” explains Martin in a new podcast produced by StateScoop and underwritten by VMware, Intel and Carahsoft.

Martin goes on to explain the difference between automating tasks and orchestrating multiple tasks or processes involving applications, web services or big data analytics.

If an agency, for example, wants to deploy a web-based application to the cloud and be able to microservice that application, it must ask what it will to look like and how it will fit into the business model. Then IT managers will be able to understand how automation and orchestration can best work for their solution.

Martin outlines several strategic steps agencies can take to benefit from automation and orchestration both for IT leaders dealing with legacy systems and those able to work with more modern systems. Among other advice, he suggests:

• Do a technology analysis of the code built into your legacy system so you understand if the frameworks are compatible with the automation toolset; and if you are using a new code, assess how compatible it will be with legacy code.
• Complete an architecture audit of the systems that will support the legacy software.
• Review your code to understand how it is written, where dependencies exist and who is supporting it. If you find code that is too old, seriously consider replacing it rather than trying to “band-aide” your tool sets or patches to make everything work.
• Look at performance to see if the legacy code will perform well in an automated task.

This podcast was produced by StateScoop and underwritten by VMware, Intel and Carahsoft.