olly - Fotolia

IT training flaws compound DevOps skills shortage

Before investing in an online course to learn new IT skills, heed this warning from brain scientists: there is a big difference between knowing what to do and knowing how to do it.

Hi, my name is Beth, and I am trained in DevOps.

Let me clarify that: as you can see by my certificate of completion, I took a 4.5 hour "Learn DevOps" course via online learning platform Udemy.com, for which I paid $10.99 (a sale price -- the regular price is $39.99). I watched -- or, most accurately, my computer played -- all 76 video segments included in the course. I also passed two multiple-choice quizzes.

My certificate of completion from one of Udemy's 'Learn DevOps' courses.
My certificate of completion from one of Udemy's 'Learn DevOps' courses.

There is no prior knowledge needed for the course, but a DevOps, cloud, Linux and networks background helps, according to the description. With 14 years of experience as an IT journalist, I have plenty of knowledge. In fact, there was little difference between my everyday work and this course – concepts were dictated, I took notes, and fed that knowledge back in to my computer via keyboard.

The course, taught by an experienced enterprise engineer, exhaustively detailed DevOps tools through lectures and demos that included Git, Ansible, Chef, Jenkins, Docker and Kubernetes. It filled in details I was shaky on previously -- such as just what the difference is between git commit and git push, and the finer points of Ansible's approach to infrastructure as code through SSH vs. Chef's client-server architecture. It also included a perfect architectural diagram of Kubernetes, which I screen-captured and saved for later reference.

But knowledge is different than know-how, and I assume perfect quiz scores don't mean an enterprise wants me anywhere near their test/dev infrastructure, let alone production. For all my informational knowledge, I was stumped by the very first question on the optional labs section found at devopschallenge.co, which even a novice sys admin could probably do in his or her sleep.

The most accessible IT training is often the least effective

Brain science research can explain what's going on here, though many of us intuitively already know it: there is a vast and important difference between knowing what to do and knowing how to do it, particularly in real-world scenarios. And the ways humans attain such knowledge are also completely separate.

"Information and knowledge is mediated by one system in the brain, [while] behavior is mediated by a completely different system in the brain," said Todd Maddox, a learning scientist and research fellow at Amalgam Insights in Arlington, Mass. "So, learning technologies that train behavior effectively will not be the same learning technologies that effectively obtain information."

Information and knowledge is mediated by one system in the brain, [while] behavior's mediated by a completely different system in the brain. So, learning technologies that train behavior effectively will not be the same learning technologies that effectively obtain information.
Todd MaddoxLearning scientist and research fellow, Amalgam Insights

Thus, while the age of the internet means e-learning sites such as Udemy, LinkedIn Learning and Skillsoft, to name a few, offer ubiquitously accessible IT training courses, there's still an IT skills shortage among practitioners who must learn how to apply information about new technologies experientially. Ironically, unfettered access to information doesn't translate into an ability to act on or apply that information.

"I still don't think we've cracked the nut on distance learning -- people are generally more geographically distributed [now], and [the industry is still] finding an efficient way to train all of that distributed workforce," said Ken France, vice president of the scaled agility practice at CPrime Inc., an Agile software development consulting firm in Foster City, Calif. "I would love to be able to find a way to fully implement and enable [IT training] through a distributed model."

The result for the IT industry is an ever-widening gap between the IT skills required by businesses that need to digitally transform -- preferably yesterday -- and the skills that are held by the pool of prospective employees, which renders many such businesses effectively unable to practically perform such a transition. This IT skills gap, along with the breakneck speed of tech innovation over the last decade, also contributes to IT employee burnout and workforce attrition, particularly in IT security.

IT training Tip #1: Take an iterative approach

Any company too small to have its own bespoke IT apprenticeship or training program -- all but the very biggest, in other words -- will encounter this problem. Experts say it also doesn't help that many enterprises take a wholesale approach to digital transformation, massive projects that attempt to migrate IT systems to public cloud and boost IT training on DevOps skills and complex cloud-native tech tools at the same time.

Even global enterprises such as UPS, however, have found an iterative approach to IT modernization most effective, and the consensus among experienced experts is the same. And for every exemplary tale such as UPS, there is also a cautionary one.

"Think about the very first year of Healthcare.gov," said Dan Tucker, vice president of digital platforms at Booz Allen Hamilton, a management consulting firm based in McLean, Va. that works with federal agencies and private sector clients. When the site first launched in 2013, it crashed and rendered health insurance marketplaces under the Affordable Care Act inaccessible -- a disaster for the federal government at the time. 

"The closer you are to the sun, so to speak, in terms of what you're delivering…doing that big bang, [is] where something like [that] could happen," Tucker said.

Many small enterprises have few good options -- they don't hire enough new employees each year to justify a comprehensive IT training program, and the kind of experiential learning they want to give employees is best served on-the-job. However, having too many trainees at the helm at once can also lead to business disaster. Companies such as Freedom Financial Network, a financial services company focused on consumer debt reduction in San Mateo, Calif., must take an incremental approach.

First, Freedom introduced AIOps automation for its service desk to free IT ops employees from repetitive daily tasks. This afforded them with the time for IT training in further infrastructure automation and DevOps technologies on the IT back end. Fourteen service desk employees will receive retraining, three at a time, in SRE skills, Tier III application support and application development.

"You can embed somebody like a junior developer into an engineering function on a product team, and get them to do smaller things like build out pub/sub utilities for moving data, or testing tools within the new QA function, to help [them] get acclimated," said Mark Tonnesen, CIO at Freedom Financial, which has an engineering staff of about 60. "But it's going to be a year, year and a half, two year journey, for them to make it to full-fledged engineer. It's not going to be a short-term process."

IT training Tip #2: Start at the top, and light a fire

People generally don't want to learn anything. They get into a place, and they want to stay there. So I think what causes change is strong leadership, a deep vision and communication about the future, and the need to update your skill sets.
Mick MillerSenior DevOps architect, KeyBank

Transforming an entire enterprise organization in product-focused BizDevOps practices must also start with a small group of people, and proceed iteratively. Here, experts recommend starting with business and IT leadership, and communicate the urgent need for digital transformation, not just to guard against business disruption, but to mitigate IT security and business resiliency risks. In other words, business leaders usually want to avoid being the next Equifax more urgently than they want to become the next Amazon or Netflix.

"Industry messaging around things like automation has been around efficiencies, i.e., [cutting] jobs, but it really should be rebranded," said Booz Allen's Tucker. "This is more about resiliency than anything else."

Clear communication about shifts in expectations can then lower individual IT employees' resistance to change.

"Training is secondary," said Mick Miller, senior DevOps architect at KeyBank, a financial services company in Cleveland. "People generally don't want to learn anything. They get into a place, and they want to stay there. So I think what causes change is strong leadership, a deep vision and communication about the future and the need to update your skill sets."

IT training Tip #3: Cultivate continuous learning

IT pros in the field echo what brain science research has concluded about the difference in value between informational and experiential learning. IT training for digital transformation must focus on the ability to learn information quickly and continuously, and then to apply that information in a hands-on process as quickly as possible.

New Light Technologies (NLT), a federal government tech consulting and managed services firm in Washington, D.C., started a weekly program of demo days, where its employees demonstrate the technologies that generate news buzz for each other and for clients, to determine how mature and practically applicable they are to real-world IT problems.

"Let's not just read about it, because we know [that] what's advertised and what is [available] are very often different," said Dave Williams, cloud architect at NLT. "We're not just going to talk about it -- we don't do PowerPoints. We show working code and working systems."

If a technology is to be put to practical use beyond a demonstration, IT training must also be repetitive, with opportunities to practice new skills over time, said Amalgam Insights' Maddox.

"This is the other thing that is completely ignored in learning and development, that the brain is hard-wired to forget," he said. "You'd be much better off training half as well, but spacing that training out to once a week … and building long term memory representation."

Dig Deeper on IT operations careers and skills

Software Quality
App Architecture
Cloud Computing
Data Center