Containerizing Legacy Apps Speeds Cloud Migration
Keeping pace with changing technology and hosting platforms is an ongoing challenge for every enterprise operation. More organizations are migrating enterprise apps to the cloud as part of data transformation strategies, but transitioning legacy apps from the data center to the cloud is never simple.
Its benefits make cloud computing extremely attractive: It’s scalable, readily available to all users (including mobile) and doesn’t require hardware maintenance. Moreover, a cloud-based ecosystem reduces costs while improving security and collaboration. But not all software can be optimized for the cloud; you have legacy software written to run in a data center and each application must be evaluated and tested before you can host them with a cloud services provider.
Running Old Software on a New Platform
To assist their customers, cloud platform providers are collaborating with customers to transition applications from data centers.
As an example, one provider that operates in over 170 countries and has a team of more than 140,000 professionals faced the challenge of assisting a client specializing in data and software services for the automotive industry with cloud migration. This automotive software services vendor processed over 300 million digital transactions annually for over 200,000 partners and customers. The goal was to migrate over 3,000 applications and retire 54 data centers in nine months.
As part of the migration strategy, the automotive vendor had over 200 software products, including an overall portfolio of over 3,000 data center applications. Many of these applications were costly to maintain and were using legacy technology stacks. They also had security and resiliency issues due to aging hardware. The entire infrastructure had to migrate to the cloud before several data center contracts expired at year’s end.
Before they could move their current software to the cloud, the automotive vendor needed to assess applications for portability. They needed a migration roadmap, a means to assess software dependencies and a way to assess the amount of effort and cost-effectiveness of redeploying each legacy application to the cloud.
Refactor or Containers?
The first step was to assess each application for cloud readiness. Rather than trying to manually validate each application, the development team opted to apply an automated software intelligence solution. The software intelligence technology rapidly analyzed each of the 3,000 applications for cloud readiness, i.e., determining what changes in the code are needed to be optimized for running in cloud environments against hundreds of “cloud blocker” patterns. It also identifies legal and security risks from open-source code components that make take a new sense of urgency once on the cloud. Using automated software intelligence made it clear which applications to keep and which should be retired.
The CIO and CTO had originally planned to refactor existing applications, rewriting the legacy code to accommodate platform-as-a-service (PaaS) once the software had been ported to the cloud. When the developer team started evaluating the software intelligence reports, it became clear that rewriting the software would be cost-prohibitive and impossible to achieve within nine months.
The only way to migrate the software and meet the deadline was to containerize the applications. By using containers, legacy software functions, including dependencies, could be standardized as a lightweight, standalone, executable software package that could run on another computing platform.
Software Intelligence Lights the Way
The automotive vendor decided to start with a subset of their application portfolio. To test the cloud-ready assessment process, they selected five of their products consisting of 134 individual applications running across nine data centers. After manually assessing only a few applications for cloud readiness, it because clear they needed an automated approach.
Using automated software intelligence, segmenting the app portfolios and creating a migration roadmap is easy. Each application was prioritized and categorized as rehost, refactor, rearchitect, rebuild or retire. Software intelligence also validated inter-app dependencies, highlighting what could break during migration. It also offered estimates of the amount of effort required to make legacy apps cloud-ready.
With the aid of software intelligence, the automotive vendor’s developer team was able to analyze 134 applications in two days, five times faster than if evaluations were done manually and with substantially higher accuracy. The insights produced by software intelligence included cloud readiness, specific blockers that needed remediation and recommendations on cloud services that each application could adopt.
Identified blockers included patterns in the code that would prohibit containerization or render it ineffective, such as local file dependencies. The relative software health of each application—its resiliency, complexity, technical debt, and agility, was also evaluated to prioritize applications and flag those that needed revisions before containerization. Software intelligence also revealed security vulnerabilities and weaknesses and applications that rely on unsecured or obsolete components.
With the aid of software intelligence, each application was assessed and classified for modification and containerization or retirement. The result was a comprehensive roadmap to show how to migrate all 3,000 applications to the cloud within nine months, including which applications would need to be retired.
Automating application migration using software intelligence proved to be the only way to port 3,000 applications and retire 54 data centers within the allotted time. Using software intelligence to deliver an objective assessment of legacy applications enabled the CIO, CTO and other stakeholders to make informed decisions without guesswork.
