The development of cloud technologies is progressing at a pace that it seems impossible to keep up with and impossible to predict. And yet, it is possible to recognize some general trends that will characterize this market in the current year, although the technologies that drive these trends, and the vendors behind them, are in a state of continuous change and are being replaced by new ones.
Many of these trends are driven by the industry entering a phase of standardization and greater interoperability, a sign of maturity in any IT sector.
Cloud infrastructure — public, private cloud hosting, and on-premise — is becoming less silo so that workloads are easier to migrate and data streams become more mobile.
This standardization – mainly through the advancement of open-source software – brings the focus to a higher level, where the new role of VARs in supporting application processes is emerging, from the introduction of artificial intelligence and high-performance computing to the provision of completely new SaaSOps services and application development.
From multi-cloud to omni-cloud
Over the past year, it has become a commonplace to say that we are moving into a multicloud world, where organizations place their workloads on an infrastructure-as-a-service from several vendors at once.
As applications become easier to port from one environment to another, computing is easier to consume in real time, data integration platforms are becoming more interoperable, and vendors are forging cross-platform alliances – this trend towards multicloud will look more and more omni-cloud. soon.
As a general rule, the largest of the organizations may soon become clients of all hyperscalers and some niche providers to boot, allowing them to receive increasingly differentiated services, special deals and avoid pegging.
The Hearst Corp., with more than 360 individual companies in various industries, is a good example of this future.
The New York-based media conglomerate recently embarked on a digital transformation with Amazon Web Services (AWS), Microsoft Azure and Google Cloud. This omni-cloud approach gives multiple corporate divisions the best competitive position in all the markets in which they operate.
Kubernetes is breaking down cloud barriers
Organizations choose the Kubernetes platform that best suits their specific operational needs and capabilities. It is often the best choice, whether it’s OpenShift from Red Hat, a stealthy implementation from Pivotal, independent distributions of similar platforms offered by Docker or Rancher Labs, or proprietary services from providers such as Google GKE, Microsoft AKS, and AWS EKS.
The container orchestrator often becomes the foundation for extending applications on disparate cloud infrastructure, providing the desired multi-cloud environment.
In this capacity, Kubernetes becomes not only a breaker of cross-cloud barriers, but also creates non-trivial market dynamics. The growing separation between the cloud infrastructure software provider and the provider who owns the server rack buildings is giving way to solutions that would have been unthinkable a few years ago.
Take Google Anthos, for example, which can run just as easily on Amazon Web Services or Microsoft Azure as it does on the Google Cloud Platform. Or VMware’s upcoming Tanzu, which goes beyond on-premises infrastructure to include all hyperscalers as well.
This multicloud world is emerging as one where not only customer workload is spread across different clouds, but cloud providers themselves boldly enter each other’s territory.
The fragmentation of related technologies
Kubernetes has won the market, but the era of unified cross-cloud that container orchestrator heralds will be complicated by the ancillary services built around it.
The Cloud Native Computing Foundation, which holds the core Kubernetes project in its hands, also harbors related technologies that will complement the platform. These future services are spelled out in the CNCF roadmap, which outlines the path to universal container adoption based on corporate use cases.
One of the technologies in these plans that is essential for organizations adopting mashups is Istio, a microservices organization tool. Other open-source projects are becoming more and more important components: Prometheus for monitoring and orchestrating container applications, ELK Stack for logging and troubleshooting (including Elasticsearch, Logstash and Kibana), Harbor – container registry for providing high availability in real environment, and Jaeger, which acts as the standard technology for tracking application logs across all microservices.
The giants of cloud services do not deviate far from the benchmark in their implementations of Kubernetes, but with regard to other developments on the CNCF roadmap, they are increasingly seeking to promote alternative technologies. For example, AWS encourages customers to use their own CloudWatch service for monitoring, rather than Prometheus. Google is doing the same with their Stackdriver.
Because this fragmentation of accompanying Kubernetes technologies will impede the achievement of cross-cloud convergence, inconsistencies may arise as more organizations move to pure cloud infrastructure.
Further Consolidation in Kubernetes …
Every cloud infrastructure provider, be it a public service provider or a vendor of locally deployed hardware and software, must have Kubernetes in their offering to stay competitive in the current market.
And once again, purchases are becoming an effective way for long-time market players to pour young blood into their newly created Kubernetes divisions or create new ones from scratch.
Among the most notable events of this kind is the mega-purchase of Red Hat by IBM for $ 34 billion, which gave it the mature Kubernetes technology in the face of OpenShift, along with the previously purchased CoreOS container pioneer.
Microsoft expanded its Kubernetes toolkit in 2017 with the purchase of Deis, and prior to that was actively involved in the project. NetApp has also entered the game with the purchase of StackPointCloud.
But perhaps the best move came from VMware with the purchase of startup Heptio, which laid the foundation for the Tanzu platform that will build Dell Technologies’ cross-cloud vision in the coming years.
But now there are even more container technology startups on the market (and Docker is one of them), so you can expect large cloud providers to actively buy them this year.
… and more shopping is safe
Many organizations, when choosing a cloud provider, want the full suite of security tools built out of the box, rather than someone else’s third-party tools.
Vendors who cannot build the security tools they need today have to buy them, and therefore, security company purchases were the leading topic in the past year.
VMware leveraged its organic growth success in cybersecurity with the purchase of client security specialist Carbon Black, and then strengthened its application security position with the purchase of Intrinsic.
Microsoft has strengthened its data protection and management position with the purchase of Blue Talon early last year. Google brought Chronicle into the Alphabet family (although some consider this not a good deal).
At the end of the year, Broadcom completed the purchase of the corporate business and the Symantec brand; Cisco acquired Duo Security a year earlier, and AT&T acquired AlienVault, making it its cybersecurity division.
But security is such a complex aspect of modern IT that there are always gaps to plug, and this buying-and-merging trend is sure to continue, if not accelerate, into 2020. There is no shortage of startups on the market that can help cloud providers of all stripes build a complete arsenal of their own security tools.
Private cloud repatriation
The inevitable shift of workloads to the public cloud is becoming more and more like a two-way street.
Containers and other technologies that make applications more portable also make it easier to fall back on a private infrastructure. Many organizations are adopting this hybrid approach to gain a better understanding of the nuances of the benefits of a given cloud environment.
This does not mean that the pace of migration to the public cloud will slow down or even wane, but that migration will be freer in both directions when customers realize that some of the workloads will bring real cost savings and increased productivity and security when hosted on a software-defined date -centers.
This private cloud repatriation will further fuel the already hot hybrid cloud market.
Everything becomes “intelligent”
Every SaaS, ITOps, analytics and BI product on offer today is equipped – in one way or another and to varying degrees – with machine learning, whether it needs it or not.
In addition to the very real benefits that artificial intelligence brings in automation and a better understanding of what is happening, the term itself serves as an excellent sales lever that it is never superfluous to insert into a product presentation.
From a chatbot to a neural network inference processor and predictive analytics, AI easily finds its way into almost any cloud software product.
Some features that use machine learning are really useful, others are just embellished with a fancy term. But in 2020, it will be difficult to find a product that is not touted as “smart.”
The growing field of SaaSOps
As SaaS continues to evolve, more specialized platforms are emerging in the market for managing migrations, IT operations and costs for these cloud applications.
In light of this, SaaSOps is gaining momentum as a new segment of responsibility for specialized IT experts – often in conjunction with offerings from a new breed of Cloud Access Security Brokers (CASB).
Services from BetterCloud, CloudManager, and Blissfully provide comprehensive management of Microsoft Office 365 and Google G Suite office suites, Salesforce.com cloud CRM applications, and other leading SaaS vendors.
“The term SaaSOps is still vague; it can be applied to security, license or expense management, shadow IT detection / reduction, and user enrollment and closure management across platforms and departments using proprietary applications,” explains Allen Falcon, CEO, Cumulus Global …
Aim at application delivery
Kubernetes has brought order to the fast-changing cloud infrastructure market in many ways. Container Orchestrator has now become something of a standard technology when deploying infrastructure to support pure cloud workloads.
Therefore, it is likely that the competition will move to a higher level, focusing on improving application delivery, said Dan Kohn, CEO of the Cloud Native Computing Foundation.
Large open-source projects will emerge to support “app dev” methods that provide services that make it easier to package and deploy applications.
In order to more successfully move towards this goal, CNCF has created a dedicated Application Delivery Special Interest Group.
HPC goes to the cloud (finally!)
High performance computing (HPC) workloads tend to run sporadically and in batch – and this is where the unique elasticity of the public cloud becomes a big advantage.
And yet, the HPC market has been driven primarily by on-premises infrastructure despite the clear advantages of the fast up and down scaling capabilities of the hyperscaler. It is only in recent years that HPC users have begun to regularly go to the cloud to remove resource bottlenecks.
This reluctance to go to the cloud has a lot to do with the nature of the work being done on HPC systems: they often help with brand new product development, proprietary data modeling, predictive and scientific modeling. Scientists, engineers and product developers don’t want their treasures to leave the walls of an organization.
However, this picture is changing, and pretty soon the expensive, multi-core systems required for this complex computing will become the prerogative of cloud providers.
The quest to stop ramping up investment in data centers while still getting guaranteed resources on demand will be unstoppable – although it is difficult to predict the price of HPC in the cloud right now.