​Cloud computing: Using containers to build vendor-agnostic approach to software development

Barry Libenson, global CIO at Experian, was handed a big task when he joined the financial data company in June 2015 -- the leadership team wanted him to drive fundamental change across the technology organisation.

Libenson describes the IT setup at Experian at the time he joined as "a bit of a free for all". While this decentralised model had worked up to this point, the senior executives running the business recognised a new approach to IT was needed to support the company's continued growth. Libenson's approach has been to embrace containers and a cloud-agnostic approach to software development.

"We are almost religious about enforcing the standards around how you build technology. That gets everyone moving in the same direction," he says.

Libenson's solution was to create standardisation in software development through containers and the cloud. "We use containers very heavily and that strategy has been something we've been working to roll out for the last two years or so," he says.

Containers decouple applications from the environment in which they run. This approach to software development means containerised apps can be deployed quickly and effectively across any environment, whether that's the cloud, the data centre or somewhere else entirely, such as an individual device.

Forrester Research suggested two years ago that 31 percent of enterprise IT organisations had deployed containers as part of their cloud operations. Recent research suggests that deployment figure could now be fast-approaching 50 percent.

In the case of Experian, Libenson has sponsored a shift towards containers through the adoption of Red Hat's OpenShift platform during the past 18 months. The firm is also using the open-source system Kubernetes to help deploy and manage its containers.

SEE: Cloud v. data center decision (ZDNet special report) | Download the report as a PDF (TechRepublic)

Libenson says software development through containers means his firm is cloud-agnostic. Rather than being tightly connected to a single provider, Experian's developers create components that can be moved between cloud platforms to meet business requirements for new services.

"So, we can build on premise, on Microsoft's cloud, on Google's cloud, on Amazon's cloud, and then we can move things around," says Libenson. "Whatever service is built must be portable across cloud providers."

Libenson says containers help to eliminate friction. He refers to the old, decentralised model of software development, where IT professionals had to pick up the phone, call someone in the data centre, and make a request for a development environment to be stood up.

Developers would then face a series of follow-up questions around the amount of memory required, processor power, storage, and the type of server they wanted to use. Two months later, developers would have what they needed -- hardware would be ordered and someone from the IT team would rack it, stack it, and install everything that was required.

"And it was painful," says Libenson. "Now, the developers don't even call -- they can go online. They can self-provision and they can stand up an environment in a matter of minutes or hours versus weeks. That has changed things, and it's all done through containers."

Libenson says Experian employs between 2,500 and 3,000 software developers globally. Today, more than 700 of these developers are using OpenShift. Libenson says this transition demonstrates how the firm has now moved from virtualisation to containers on a broad basis.

Kubernetes provides orchestration, so developers can move the containers between different cloud platforms. This containerisation and orchestration allows developers to create services on any cloud platform quickly and securely, says Libenson.

"When a software developer at Experian sits down to build a new product, they may not even know whether they're building it in the US, the UK, in Microsoft's cloud, Google's cloud or on premise. It doesn't matter -- they see the exact same thing. They see an identical development platform, regardless of what data centre they're building out of," he says.

Experian's customers are typically large financial institutions that are looking for information for campaigns. Other key clients include automotive firms and healthcare organisations. Libenson gives the example of a client that might want to know how many individuals in a specific geography have a high credit rating.

"Our tools allow them to get the insights that they've never been able to have before to figure out whether they should be lending more cash or less. There's an enormous amount of information in our system -- and it's lightning fast. It's built on a very leading edge set of technologies to provide that capability," he says.

Libenson says everything Experian builds is designed to be hybrid in nature. That way services can scale dynamically in response to seasonal trends. The firm tends to see a big increase in activity from clients around the holiday season, particularly when it comes to credit card applications and fraud detection services.

SEE: Sensor'd enterprise: IoT, ML, and big data (ZDNet special report) | Download the report as a PDF (TechRepublic)

The combination of containerisation and orchestration allows Experian to cope with these seasonal trends. The firm has created a flexible, cloud-based platform that can be scaled upwards or downwards without huge expense. Libenson advises other CIOs to think about their demands and to consider whether they would be better-served using the cloud and containers.

"Do you build the church for Easter Sunday or do you build the church for normal Sunday worship?" he asks. "If you build it for Easter Sunday, you get this giant structure that must exist every day of the year, but that only gets used one day of the year. So, you have a lot more capacity than you generally need, and there's a really high cost associated with that."

Experian has supported containerisation with a move to Agile development practices. Libenson says this move has entailed a pretty big shift for the IT organisation. Developers can now create and change services far quicker than under the traditional model of enterprise IT.

Gone are the days of the old enterprise software model, where once a year the IT team is notified of a new release or maybe once every 18 months, says Libenson. IT directors and their line-of-business colleagues today face a much more frequent pace of change -- and Agile, containers and the cloud are set up for that kind of reality.

"There's an expectation now around applications and the frequency at which they get updated," says Libenson. "That expectation requires that you do Agile and roll out new features and new functions pretty much monthly."

RECENT AND RELATED COVERAGE

Cloud computing: Three reasons why it could be time to go cloud-first

Why are more CIOs putting on-demand computing at the top of their to-do list?

How to use cloud computing and big data to support digital transformation

Parkinson's UK is using cloud-based data warehousing to support changes that put better use of information at the heart of its processes.

NFL in London: How big data is helping to change the game when it comes to player performance

New technologies are giving NFL coaches, and football fans, better insights into how players are performing.

Five ways to talk about tech, without switching people off

Talking about why tech matters can be hard. Here are some tips on getting your message across.

Why today's containers and microservices will be tomorrow's legacy sooner than you think (TechRepublic)

The pace of software innovation is increasing, as is the pace at which your cool code will become legacy tech.

Adobe's Moving Stills tech turns a regular photo into a mini-movie (CNET)

Adobe Max sneak peeks show AI and automation for font styling, video cropping, illustration morphing and video subject selection.