If there are two departments that have undergone huge changes in recent years, it’s development and operations, and 2017 looks likely to see this trend continue. What’s on the horizon for the often dev-centric DevOps teams and operations in particular?
A test of faith for DevOps
According to Gartner Research, DevOps is at the peak of inflated expectations and staring at the trough of disillusionment? Is this correct? During the Gartner Data Center London event last month, it was revealed that 38% of Gartner Circle members stated that they were already using DevOps today but equally, as presented at the aforementioned event, Gartner conferences witnessed 87% of attendees stating that DevOps has not delivered the benefits they were expecting. “Changing the behaviors and culture are fundamental to the success of a bimodal IT approach. We estimate that, by 2018, 90 percent of I&O organizations attempting to use DevOps without specifically addressing their cultural foundations will fail,” according to Gartner Research Director Ian Head. So, a lot of buzz surrounds the topic, which at my count earlier this month returned nearly 17.5 million search results.
Making the move to a DevOps approach is not for the faint-hearted or easily discouraged. It takes persistence, belief, and superior internal sales skills to lead others on the journey. The good news is there is now a critical mass of enterprises who have made the move and are experiencing significant benefits. It just needs a tactical approach to advocate and oversee change in the face of opposition, momentum-sapping inertia, or difficulty adapting. In doing so, initial wins can be achieved, upon which further initiatives can be built.
An equal partner with the business
As DevOps makes its impact (high-performing IT organizations deploy 200 times more frequently than low performers, with 2,555 times faster lead times for example), the relationship between IT and the business becomes intrinsically interlinked. The capabilities which deliver better quality, more robust applications faster, and with less waste open up significant potential for new customer offerings, improved customer relationships, and time-to-market. In short, DevOps adoption can mean a critical competitive edge. The decision around what to build, when to release it, and when to update it should be the result of an ongoing peer to peer dialogue between tech and business leads. In parallel, IT teams overall need to position themselves as enablers of transformation, not inhibitors.
A few years ago, any DevOps introduction would almost inevitably include a classic “silo” picture of the dev team on one side and the ops folks on the other (and yes, I was one of those offenders). This situation is evolving now as new roles that blur this division emerge such as DevOps Engineer, Site Reliability Engineer, and Cloud Architect. They don’t sit easily 100% in one camp or another and possess hybrid skills that can prove a real asset in delivering robust, scalable, and rapidly deployable applications. Expect new structures to emerge that resemble a flotilla of speedboats rather than an ocean liner in terms of ability to respond to changing demands.
No metric blind spots
In an increasingly data-driven world, the complexity of today’s applications should not be an excuse for the unavailability of insights into how they are performing. Whether or not an organization has chosen public, private, or hybrid cloud, experimenting with microservices or embracing the potential of containers, rich, detailed, yet easy to understand metrics around every aspect should be at hand in real-time.
IoT and DevOps: a brave new world
The IoT topic has been around for several years, but 2017 could be the inflection point. Gartner Research estimates that by 2020, more than half of major new business processes and systems will incorporate some element of the Internet of Things. While wearables may have the consumer eye, industrial IoT usage dwarfs that in consumer markets. As a counterpoint to industries that have led in digital transformation (retail, banking, and telecommunications), the heaviest IoT users are likely to be in oil, gas, utilities, and manufacturing industries, according to a global survey released March 3 by Gartner. No sector is immune to the need to review and evolve its application development approaches. In parallel, the DevOps relationship to IoT is an interesting one, particularly around the end user experience (it’s a myriad of small devices now, not a tablet or smartphone) and security (think public wifi issues) to name but a few. Expect more DevOps teams to work on applications with an IoT use case and go through a major learning curve.
DevOps and its relationship with data
It’s not just big data, all data, including databases themselves (which came up often at AppSphere 2016 as a cause for latency and downtime) that will matter. From the rise of the role of data scientist to the explosion of IoT data, DevOps teams cannot ignore this all too important area, and need a POV regarding how it should be managed. One particularly interesting area will be the increased use of business algorithms, with a lot of the data needed to build these held within the remit of IT operations teams. Machine learning APIs can have a significant role here, as they help developers to apply machine learning to a dataset enabling predictive features to be added to the applications they are building. One example of this is Google Prediction API, a cloud-based machine learning and pattern matching tool. It helps with the upsell of opportunity analysis, provides details of customer sentiment and churn analysis, and detects spam, amongst many other features. Stephen Thair of DevOpsGuys has a written a solid exploratory piece on this topic, which may not be as hot as containers, but is still an essential consideration. Data Ops experts share a related goal (faster time-to-market) and DevOps centric organizations would be wise to have a line of communication between the two teams.
This subject could be a blog all on its own. How can DevOps work more closely with security teams as data breaches threaten to damage a brand as much as a slow responding app? What can DevOps do to ensure applications are built with security in mind from day one? How can anomalies or outliers in business transactions provide an early warning system to fraud? The potential for APM insights to assist in fraud detection is an interesting use case which PayU, one of our customers, raised in their presentation at AppSphere 2016. Look out for more data breach headlines in 2017, though maybe not at the scale of the recently announced Yahoo hack, which is understood to have involved one billion accounts.
But what about operations, specifically? In some ways, the less prominent player in DevOps adoption, ops teams are perhaps more likely to undergo even greater changes than their dev counterparts in 2017.
Move from being a cost center to an innovation hub
Enterprise technology spending in 2017 is set to rise 3%, with $3.5 trillion on technology expected to be invested, according to Gartner Research. Yet within this environment, operations leads will still need to pivot from being seen as an area that risks being cut year-on-year to one that helps build the business and is in step with its goals. This requires many behavioral shifts — from running a tight ship to being a negotiator and intermediary between multiple cloud, software, and infrastructure vendors, and lines of business themselves. The ability to market one’s team and its contributions becomes almost as important as deep technical knowledge when pursuing investment. The established focus on being the data center guardian and only objectivized on stability has to shift towards accepting that the I&O has to be anti-fragile, rather than simply unchanging. This will take a fine balance between calculated gambles on new approaches and technologies, and acceptance of the fact that, as Ian Head of Gartner Research pointed out, a small amount of risk has to be accepted as it is impossible to innovate otherwise.
No ops team is an island
Having just misquoted John Donne, those in operations will need to accept that they will both lose and gain going forward. As Gartner Research pointed out at the Data Center Summit, infrastructure and operations environments have traditionally been the center of the ops team’s world view. However, as the organization expands, operations staff will find that while the number of activities they become involved with increases, their ultimate control over specific areas reduces at the same time. Organizations are now taking a more collaborative approach. This means the creation of flexible and agile networks and ecosystems become ever more important, with the innate capability to scale and rollback investment areas as the business demands. This requires an astute ambassador with a business-savvy mind, building agility into the I&O mindset rather than a rigid enforcer who views change with suspicion.
Customer experience is key
As DevOps delivers an improved customer experience, there is potential for more insights as to exactly what the end user sees and interacts with, bringing development in particular closer to understanding exactly whether what they have built works, or doesn’t, with its target audience. At AppDynamics, we have seen how end user monitoring of experiences across mobile, tablet, and PC platforms and devices, for example, is essential for understanding how, when, and where customers are engaging with an application. This has been critical in preventing a degradation in response times ahead of impacting customers and the company’s reputation.
The skills gap
How do enterprises build and nurture teams that are equipped for the digital business platform? Relying on rockstars and contractors are short-term fixes. How can the classic Gartner Research I&O employees working on predominantly Mode 1 refresh their skill sets and feel part of the “new way”? How can up and coming young talent be attracted not just by salaries, but a culture where they feel valued and listened to? DevOps can’t be just for the rockstars and stellar contractors — buy-in needs to come from colleagues who see the move as inclusive and non-elitist. There are many external consultancies who can provide some excellent stewardship and direction in the tricky subject of enterprise DevOps, but the best ones are those who teach others how to fish, pair up, and coach them — rather than treating knowledge transfer as an inhibitor to additional consulting fees.
If there is one standout theme for 2017, it’s the overwhelming need to revisit how dev and ops view their role in the organization, how they contribute value, their scope of responsibility, and the new mindsets needed to thrive in an ultra-high velocity world. In the words of Stephen Covey, continually “sharpen the saw”.
Be hungry to learn, question perceived wisdom if it doesn’t fit with evolving demands, and remain open to trying new approaches without fear of failure inhibiting new platform initiatives, onboarding technology partners, and being objectivized on more outcome-based metrics.
Download the eBook, 10 Things Your CIO Should Know About DevOps.