Warakorn - Fotolia

IT ops will reap the benefits of API economy, too

An understanding of APIs is table stakes for IT operations pros in the modern job market, as businesses increasingly operate and connect through these interfaces.

Resistance to the API economy is futile -- even IT operations will be assimilated.

APIs began as the domain of developers, but the benefits of API use also include the ability to automate IT incident resolution and integrate ops management products even where vendors don't offer the integrations natively. Even when IT operations' strategy doesn't directly require APIs, IT operations should monitor and keep them reliable, as companies increasingly offer them to partners and customers, according to enterprise IT pros.

APIs now dominate both business and IT -- a phenomenon market analysts have dubbed the API economy -- thanks to technical trends over the last two decades, from the proliferation of the web to Moore's Law, which have condensed powerful computers into ever-smaller packages that are well-suited to microservices.

Meanwhile, with the web and standard API specifications such as REST, programs no longer require deep knowledge of where nodes reside on a network to communicate between disparate compute services.

"You don't have to know nearly as much today as you had to know 10 years ago to build [microservices-based] applications, and that's why almost everything you see out there is moving in this direction," said Barry Libenson, CIO of Dublin-based credit bureau Experian, where APIs have changed how business is done and how IT operates in recent years.

The API economy has also been created by cultural change in the business world, Libenson said.

"People don't like having the way they consume information dictated to them," he said. "They want to be able to write applications or create programs and then pull data in real time, and they want only the data that they need in order to act."

The benefits of API deployment for app developers are easy to identify. They allow for communication between the open source components that have become increasingly popular as building blocks for modern applications and can be used to automate DevOps pipelines.

Meanwhile, IT infrastructure has embraced open source building blocks, as well, and APIs present several benefits for IT operations teams.

DIY integration of IT operations tools

Let's face it: Vendors don't always make tools that play nicely with others, or with features that address every corner case. That's where the benefits of API expertise come in for shops such as Jamf, a mobile device management company in Minneapolis, which uses REST APIs offered by collaboration tool-maker Atlassian to create its own product integrations and customize app deployment workflows.

Six months ago, the DevOps team at Jamf initiated webhooks and API connections between Atlassian's JIRA, Bamboo and Bitbucket Server, so when an engineering team puts in a request to do a release, it appears in JIRA Service Desk, creates a JIRA card on a JIRA Kanban board, and DevOps engineers can move it from column to column on that board. Under the covers, as the card is moved, API integrations make calls to Bitbucket Server and Bamboo to create code branches, create builds and do the deployment.

"We call it 'release from your iPhone,'" said Michael Kren, manager of DevOps at Jamf. "We can kick off a release just by moving a card in JIRA."

Jamf's DevOps team has a motto that they're going to automate themselves out of a job, Kren said.

"And that's OK -- then, you get something more fun to do," he said. "Now, we can focus on, 'How do we containerize our products so we can run on any set of infrastructure? Or, how do we allow a nontechnical individual to spin up a test environment with the push of a button?'"

API benefits such as deployment automation mean Jamf's DevOps team has held steady at three employees, while the engineering team has doubled in size in the last year.

"We've built these things using the APIs to take care of housekeeping, so we can concentrate on moving engineering forward," Kren said. "If you keep yourself ahead of engineering, you can find yourself suddenly adding tremendous value to your organization."

Automated troubleshooting

If you keep yourself ahead of engineering, you can find yourself suddenly adding tremendous value to your organization.
Michael Krenmanager of DevOps, Jamf

When apps and IT operations tools are joined together through APIs, IT incident resolution and infrastructure management can be streamlined.

At Lineage Logistics -- the second-largest warehousing and distribution company in the world, with over 100 U.S. locations and some 10,000 household-name food manufacturing customers -- the warehouse management system's (WMS) availability is the primary responsibility of IT operations.

An API-based integration and custom XML-based modules integrate the company's homegrown WMS with Kaseya's Virtual System Administrator product. IT ops uses this system to automatically and continuously check the health of the WMS, and flag issues for ops personnel to investigate.

"We use Kaseya to go out and verify transactions within our WMS system to make sure it's actually doing what it's supposed to do," said Buster Davis, technology manager for the Irvine, Calif., company's southeast region in Georgia.

"Eventually, we will roll it into doing more automated remediation, but we want to see from a development standpoint early on what are the issues that are recurring, and what can we do to minimize them?" Davis said.

A few months ago, the company still relied on human intervention to check systems and notify operators, a process which was slower and more error-prone. Dated software modules within the WMS alerted IT ops to problems, which also required more human intervention than the Kaseya system, Davis said.

The API imperative

Even if IT operations pros don't write their own APIs or interact with them directly, lines of business will use APIs to exchange information with business partners and provide services to customers. Thus, IT pros must understand the many-layered nature of API infrastructure.

"As business services, infrastructure and operations teams need to be able to talk about, monitor and know that APIs are being delivered in a bulletproof fashion," said Randy Heffner, an analyst with Forrester Research. "They need to be prepared to very rapidly do root-cause analysis."

At the edge of the enterprise network, there might be an API gateway that integrates into an infrastructure platform, such as an enterprise service bus, or the gateway may call an application, which calls another application, which calls a database and so on.

When something goes bump in the night, you could call all the teams associated with all those layers, or your tooling can be set so it connects the dots and identifies, 'It's the third layer down.'
Randy Heffneranalyst, Forrester Research

"When something goes bump in the night, you could call all the teams associated with all those layers, or your tooling can be set so it connects the dots and identifies, 'It's the third layer down,'" Heffner said.

"The other tool you might have in your toolbox is the ability to dynamically run synthetic transactions against APIs to find out if they're working or not," Heffner added.

Developers who create APIs should also deliver test harnesses that can run synthetic transactions at least during regression testing, if not production, he said. Tools such as Software AG's WebMethods Insight and Dynatrace's monitoring platform can also come in handy for ops pros to monitor APIs.

Beth Pariseau is senior news writer for TechTarget's Data Center and Virtualization Media Group. Write to her at [email protected] or follow @PariseauTT on Twitter.

Next Steps

APIs prove they have a broad scope of use

Consider the API when choosing a SaaS

Using APIs lets vendors convert hardware to software services

Dig Deeper on API Management for IT Tools