IT infrastructure teams face a DevOps and cloud-native technology skills gap

November 11 2019
by Henry Baltazar, Liam Rogers


Enterprise infrastructure teams are facing many challenges, the amount of data under management is climbing, the amount of work they are taking on is increasing, and team sizes remain stagnant at many organizations. To further complicate matters, emerging technologies can put infrastructure personnel at a disadvantage as their organizations assess their existing skills gap and how best to combat them. Our Storage, Organizational Dynamics 2019 survey shows that many organizations are contending with such skills gaps, and DevOps skillsets are a leading shortcoming.

The 451 Take

As more organizations tread down the path toward hybrid and multi-cloud, as well as cloud-native application development, their infrastructure teams will need to master an ever-broadening range of technologies. Implementing DevOps-related tooling and methodologies will necessitate that infrastructure teams shift some of their focus from core infrastructure skills like server and storage administration and more toward areas of expertise such as container administration. Even as tools like Kubernetes gain more traction in the enterprise and infrastructure and DevOps personnel become more adept at leveraging it to reduce the frequency of manual tasks via dynamic and self-service provisioning, there will still be cultural and trust issues as well as a desire to have staff oversight – this is reflected in our interviews with end users.

To close skills gaps, training existing employees and hiring new ones are the predominant options and have grown more common year over year. While using third parties like service providers and contractors remains a viable option, many organizations are eschewing them in favor of enhancing their current staff's abilities and expanding their teams with new hires.

Skills gaps

For 2019, 52% of organizations indicate that they are currently experiencing a skills shortage among their infrastructure-based personnel (See Figure 1). This is an increase from 2018, when the percentage was 42%. The number is even higher for larger organizations, those with 1,000 or more employees, where 60% are experiencing a skills gap. We see that organizations with a preference for generalists when hiring new candidates are less likely to be experiencing a skills shortage than those that prefer specialists. Sixty percent of organizations that prefer a generalist approach have no skills shortage, compared with 42% of those that prefer specialists. Organizations that prioritize hiring specialists may have a harder time filling roles, while those that prefer generalists are more likely to have reduced expectations, especially if the current environment is architected to be generalist-friendly or their intent is to train them on new skills later.

Across the board, we note fairly similar numbers between those that prefer generalists (29%) and those that prefer specialists (27%) – the remaining 44% currently prefer an equal mix. Those that prefer generalists cite the primary reason (68%) as a desire for a lean and broadly capable IT team, indicating that the driving force is not the high cost or difficulty in finding specialists (3% and 1%, respectively), but rather the desire to have personnel with a range of skills.

Figure 1
Figure 1: Companies that prefer specialists experience skills gaps the most
Source: 451 Research's Storage, Organizational Dynamics 2019

However, we also see that the areas of expertise where skills gaps are occurring are not those related to traditional core infrastructure skills. The areas where skills are in short supply include DevOps (47%), container administration (39%), database administration (35%) and data science (33%) (see Figure 2). These categories have crept up since 2018, underscoring the growing importance of these fields and their related technologies.

Additionally, we note that core infrastructure skills have decreased more significantly over the past year, with gaps in server/system administration-related skills falling from 43% to 32% and storage admin dropping from 33% to 26%. DevOps is having a direct impact on infrastructure teams given that it requires participation from operations and not just developers. Organizations that want to incorporate DevOps will need to modify infrastructure skillsets to be compatible. There is a similar dynamic with cloud-native technologies such as containers and Kubernetes. These technologies and their usage can span multiple disciplines and will require some buy-in or support from infrastructure teams, so we expect that skills pertaining to container administration and orchestration will be an expanding area of interest.

The value of data is a message that has been driven home for years and the desire for data scientists and database admins will increase as long as data under management continues to climb rapidly and organizations recognize the importance in leveraging it to create business value through analytics and machine learning. While some platforms enable easier provisioning of infrastructure resources like GPUs for machine learning and artificial intelligence (AI) use cases, supporting them still requires some administration on the part of infrastructure teams.

Figure 2
Figure 2: Cloud-native skills are in demand
Source: 451 Research's Storage, Organizational Dynamics 2019

Homing in on DevOps, the narratives below from our recent Storage user interviews demonstrate the perceptions some organizations have regarding the topic. Some customers see the potential benefits – however, incorporating a DevOps approach within an enterprise can be challenging. Not only does it necessitate some reshaping of methodologies and tooling but it can also require cultural changes. Infrastructure personnel will have to adapt to help better align the priorities of stakeholders and individual teams. Our DevOps 1H 2019 survey elaborates on the benefits of DevOps, including cultural synergy, as well as the cultural challenges that are being encountered with the hurdle of overcoming resistance to change and leading the pack.

'There is a DevOps team. There's conversations. They've been around to say, 'How can we help you?' And we haven't found that they have enough technical knowledge yet to help us on the storage side but those discussions exist and there's discussions that maybe eventually we'll take a storage person and move them onto the DevOps team to see how they can help.'

IT/Engineering Managers and Staff, Financial Services

50,000-99,999 Employees, $10bn+ Revenue

'[DevOps] is integrated within our environment because, without it, we will not be able to move fast enough.. [Usage has been] about two years now, heavily.. The teams didn't have to grow and there was a little bit of reshuffle between the teams but overall, [DevOps] has just improved the efficiency.'

IT/Engineering Managers and Staff, Consumer Retail Products & Services

100,000+ Employees, $10bn+ Revenue

'[DevOps] is a cultural shift that we have to make, and then you apply the methodology. But if you apply methodology but you still think in an older way, it doesn't really work. So that's really the cultural transformation in many companies.'

Mid-Level Management, Manufacturing

5,000-9,999 Employees, $5-$9.99bn Revenue

Most organizations plan to train existing staff

The most prevalent choice for addressing skills gaps is to train existing staff (see Figure 3). However, 52% of organizations plan to higher new personnel to close skills gaps, and this option has increased 10% year over year. At 59%, larger organizations (in this case 1,000+ employees) are also considerably more likely to turn to hiring existing staff compared with smaller organizations (those with 1-999 employees) at 46%. Larger companies may be in a better position to take on additional staff, but organizations of all sizes still favor training existing employees: 62% of smaller businesses plan to do this compared with 68% of larger organizations. Turning to third parties in the form of contractors, service providers and outsourcers are the less-favored options.

Figure 3
Figure 3: Training up current staff is the leading method for coping with a skills gap
Source: 451 Research's Storage, Organizational Dynamics 2019

The following narratives illustrate the types of drivers that motivate organizations to train their existing employees. These drivers include the need for more coding ability in modern infrastructure management and automation. However, as the third narrative highlights, some organizations will turn to training existing staff because they encounter difficulty hiring.

'Everything nowadays is PowerShell, and you need to know a little bit more about how to code the infrastructure. So infrastructure is code versus infrastructure is hardware. So that's where we're seeing shifts.. Some of us older guys are like, 'Okay. Well, we'll bring in some new people to do that, and we'll supervise.'

IT/Engineering Managers and Staff, Financial Services

100,000+ Employees, $10bn+ Revenue

'We're just in the process of adding a bunch of automation and orchestration.. I think it will require people to learn a different skill, and how to automate, and how to monitor better. I don't think it's just going to be, 'Hey, we've automated all this stuff, we can get rid of half our people.' You need people to be able to manage the automation, and ensure that it's working, and continue to build it out.'

IT/Engineering Managers and Staff, Financial Services

100,000+ Employees, $10bn+ Revenue

'We're looking for folks with that type of expertise [converged and hyperconverged infrastructure]. And we've had difficulty finding them.. What we tried to do a couple of years ago is to hire lesser-experienced folks, and to train them up in this technology. But we haven't been truly successful.. It's been taking too long, and we don't have a good enough training platform yet, or a training infrastructure in place to tolerate the time and effort that it takes to properly educate the new folks.'

Senior Management, Financial Services

10,000-49,999 Employees, $5-$9.99bn Revenue

The increased momentum for training existing employees can be attributed in part to challenges with recruiting. In our Storage, Organizational Dynamics 2018 survey, we saw that 82% of organizations were encountering some difficulty in recruiting for infrastructure-based roles. Top reasons for this difficulty included a lack of skills and expertise in candidates, as well as a lack of viable candidates in the organization's geographic region and salary-asking prices that were too high. Training existing staff is one remedy for those unable to hire new personnel, but it is not without its own challenges since it can be a gradual and time-consuming process. Ultimately, infrastructure modernization will require making an investment not only in technology but also in staff. Some employees are also reluctant to get automation training for fear of being automated out of a job, so management needs to let infrastructure professionals know that automation skills will benefit them by extending and potentially transforming their careers. Automation and management tools augmented with machine learning have great potential to reduce the workload on IT teams, but they are not silver bullets, and for the foreseeable future organizations will require specialist and generalist staff that can oversee infrastructure.