Looking ahead to 2020 – Joel Friedman, Rackspace CTO, offers his predictions for the upcoming year
As we enter 2020, hybrid and multi-cloud will continue to gain traction, as will software as a service (SaaS). Security remains a tough nut to crack, and edge computing will gain ground, but I predict more modestly than the current hype may suggest.
Prediction 1: Hybrid/multi-cloud remains ascendant
Smart organizations have locked onto the benefits of running apps and data in the places that make the most sense, whether that’s public or private cloud, colocation or, most commonly, a bespoke combination of these that can evolve as business needs and resources dictate. However, the complexity of managing multiple clouds across the dimensions of cost, security, governance, identity and DevOps patterns will continue to grow in 2020.
This complexity is related to the forced reliance upon a multitude of platforms, the rate of technological change, disruptions to traditional service delivery models and a real shortage of skill sets required to bring it all together. Because of these factors, it is likely that multi-cloud solutions will prevail, while the search for qualified external help will intensify.
Finding external help may prove difficult, as not only will managed service providers need to maintain a multi- or hybrid cloud instance, but internal and external vendors will need to work together to manage a multitude of issues around cost governance, security and tagging, to name a few. Fortunately, 451 Research has found these growing pains can be alleviated with clear communication and flexibility on each partner’s part.
Prediction 2: SaaS = problem solved
On top of infrastructure, one of the cloud strategies I hear on repeat when working with customers is that “cloud first” starts with SaaS. They don't want to reinvent the (inferior) wheel by developing applications for 'solved' problems. Mature SaaS offerings are a culmination of many iterations of customer mandated features, user experience analysis, and just as important, industry best practices around process workflows.
In many cases, I see organizations choosing to fit their internal process to best match their SaaS vendor’s implementation – as opposed to always opting for customizing the platform to suit legacy workflows. Furthermore, SaaS has a compounding effect with data; once data is in the system, other ecosystem players can offer turn-key integration and add-on services, further enhancing the value proposition of SaaS. While this isn’t necessarily a new trend, I predict this will be the norm in 2020 and I expect to see more and more niche and vertically-focused applications developing a SaaS model.
Prediction 3: Security remains a struggle
When it comes to security, many organizations remain burdened with legacy systems riddled with technical debt and corresponding security vulnerabilities. They understand the need to harness cloud-compatible security frameworks, operating models and tools to operate securely in cloud-native environments, but it can be slow and rough going.
Most industries simply have not reached the place where mature cloud security practices are being implemented, and this creates fertile ground for breaches. Too many are still attempting to apply traditional security controls and methodologies to cloud-native environments and deployments. This generally does not work well; not only is the security implementation likely to be ill fit, it also creates drag on the organization's desired benefits of agility. This sometimes leads to fragmentation, where business units go around central security and implement on their own shadow security. The risks of this should be obvious.
In 2020, I think the cloud continues to present some growing pains to even the most digital-native and forward-thinking organizations. These organizations understand how the cloud can aid their business in moving fast, but they don’t yet have the maturity to implement real-time or just-in-time posture management and ‘least privilege’ protocols in the ephemeral, complex and constantly changing world of multi-cloud. While security teams straddle the old world of insecure legacy applications and platforms (those which cannot or should not be modernized), and the relative newness of cloud operating models, I unfortunately expect to see the breach headlines continue in 2020 along with all of the free credit monitoring we can handle.
Even when the security teams agree in concept that it’s possible to be more secure in the cloud, many do not have cloud compatible security frameworks, operating models and tools to aid the business in operating in cloud-native environments securely.
Prediction 4: Cloud control plane will become the norm
There has been a trend over the past few years in which the management plane has been reversing from the datacenter to the cloud. For early cloud adopters, the datacenter was still the central control point. Organizations burst into the cloud or had back-end projects with no internet accessibility. As cloud grew in popularity, more and more greenfield projects became cloud-native.
Granted, many still integrate with on-premises or hosted private clouds for identity, business intelligence or other data enrichment requirements, but the hyperscalers have now moved past denying that hybrid cloud is real (in attempts to capture all workloads), and pivoted their strategy to capture those datacenter workloads where they stand and bring them into their ecosystem. This includes Snowball Edge, AWS RDS of VMware, Azure Stack/Azure Arc and Google Anthos. Expect to see more such services in 2020. And why shouldn’t we? Cloud providers have proven their effectiveness of securely operating at scale, and API-enabling everything.
Prediction 5: Edge will gain moderate ground
Edge computing is a new frontier — no player is currently dominating this space, as it is a naturally fragmented market. When choosing locations over platforms to address local markets, data sovereignty, and other use case specific needs, I believe organizations may want to align in a technology-neutral and consistent manner. And based on the trends over the past year, containers and serverless seem like a natural beneficiary for edge.
And while Kubernetes dominates in the container orchestration arena, serverless is another story in which, as of yet, there are no victors. AWS, Azure and Google Cloud Platform all have their own flavors of Function-as-a-Service (FaaS), which are embedded within their respective cloud ecosystems. Will OpenFaaS with kNative gain some ground based on open standards, addressing the often spoken lock-in avoidance and mobility potential? We shall see. Either way, I predict that 2020 will see lots of innovation and exploration in the edge space, but I suspect this is the year of gaining momentum and the floodgates won’t open until 2021.