Sign up for our newsletter and get the latest big data news and analysis.

2021 Trends in Cloud Computing: The Omni Present Multi-Cloud Phenomenon

Cloud computing has become as indispensable to the data ecosystem as it has to consumers’ lives. Its collaboration capabilities (eminently evinced in telemedicine, online banking, and e-commerce) epitomize the remote interactivity of a big data landscape shifting, inexorably perhaps, towards the edge.

Less than a year ago, this decentralized transaction capacity was a matter of competitive advantage. Today, it’s become the cloud’s cardinal point of distinction as its architecture keeps society itself operational—and connected.

As such, it’s imperative organizations optimize its traditional advantages of reduced cost, scalability, and ubiquity of access with a low latent experience. Although numerous developments have arisen to reinforce these boons, their underlying motif is the multi-cloud phenomenon surging to the top of progressive organizations’ priorities.

“If you look at Gartner, they say 80 percent of customers will be using multi-cloud in the near future because each of the cloud service providers have different strengths in their technology stack,” reflected Denodo CMO Ravi Shankar.

Multi-cloud deployments are central to cloud computing for numerous reasons, the most immediate of which may be an expansion of viable public clouds. Although Google (prized for its Artificial Intelligence utility), Azure (lauded for workplace applications), and AWS (desired for its B2C accessibility for retailers) remain the most prominent, clouds by Oracle and IBM are gaining credence too.

Moreover, the cloud’s pervasiveness throughout personal and professional spheres of life warrants multiple clouds, types of clouds, and support for distributed technologies. According to Jacob Smith, Equinix VP of Bare Metal Strategy & Marketing, “The compelling thing is like, I’m BMW and I want to deliver mobility, and that’s going to require me to be everywhere, and everywhere looks like a lot of different things. That’s what’s driving this idea of cloud everywhere and what people want out of that: automation everywhere.”

Automating the multi-cloud experience to profit on its aforementioned advantages requires culling the best opportunities for data integration, cloud native technologies, data privacy, private clouds, and security.

Organizations capitalizing on multi-cloud deployments can not only standup instances anywhere on demand, but more importantly, realize this architecture’s full potential by “doing multi-cloud integration as a Swiss Army knife going across multiple different clouds without having to be horned into a particular cloud,” Shankar affirmed.

The Data Fabric Presence

The data fabric tenet directly addresses the automation requirements of a multi-cloud reality by seamlessly integrating data across clouds (and other environs) for centralized access, analysis, and action. Despite the abundance of implementation methods for these platforms, a logical data fabric offers “a real-time integrated view of the data, whether it’s structured, unstructured, cloud, on-premises, data at rest, or data in motion, and then publishes it to the consumer without having to replicate the data to a repository,” Shankar noted. This approach leaves data wherever they are, yet provides uniform access to them for timely action. Advantages include a universal semantic layer for redressing schema and terminology differences, publishing to any tool or cloud of choice, and deriving structure from semi-structured or unstructured text.

Data fabrics also directly redress the varying forms of vendor lock-in that are antipodal to multi-cloud functionality. Several cloud “vendors are looking at their ecosystem: they want people to ingest as much data as possible within their systems and penalize them for taking the data out,” Shankar cautioned. The rapidity of cloud data warehouses and variety of data structures they encompass are meaningless if their data aren’t shareable. Logical data fabrics supersede such limitations by “one, reducing the egress charges, the time to data; you don’t have to replicate data across multiple clouds, and [they] also provide that uniformity and security across all these,” Shankar posited.

Cloud Native

Cloud native technologies are the fundamental enabler of the multi-cloud experience because of their overarching portability. They support a best-in-breed deployment approach and situations where “say I’m a bank and I need to be in Geneva because regulatory compliance says I have to be there,” Smith propounded. “So, suddenly you’re Bank of America and you’ve got 50 different regulated regions. You have no choice but to be multi-cloud.” The dynamic mobility of cloud native capabilities partially obsoletes the hybrid cloud term—applicable to combinations of public and private clouds and on-premise deployments—since “multi-cloud is just a more modern term of hybrid cloud,” Replex CEO Patrick Kirchhoff specified. Cloud native automates multi-cloud experiences with:

  • Infrastructure as Code: IaC transforms the notion of infrastructure from physical hardware to layers of software in cloud settings. This transition heralds the fact that now “software is the consumer of infrastructure,” Smith stipulated. IaC’s typically implemented by visual, low code approaches (via drawings) that generate website or application code. When coupled with other cloud native methods involving containers and orchestration platforms “we can both shift features and scale everyday, and be stable,” Smith remarked.
  • Containerization: Gartner projects worldwide container management revenues to reach approximately $950 million in 2024. These lightweight repositories are frequently the most utilized cloud native capabilities because they transport workloads, data, applications (including databases), and microservices wherever they’re needed: between clouds, at the edge, and on-premise. Containers are also critical for runtime orchestration; Kubermatic CEO Sebastian Scheele indicated that, “Containers, at the end, is a way to start processes and manage processes in an isolated way.”
  • Orchestration Platforms: Automation platforms for container management solutions like Kubernetes mask the complexity of operating disparate containers at scale “to run on different cloud providers and on-prem hybrid setups that bring flexibility, as more and more enterprises adopt this multi-cloud approach,” Scheele acknowledged. Automating these tools is critical for “the whole familiarization around container native functions, where the main purpose is, in the future, running the whole network for 5G,” Scheele remarked. Emergent Kubernetes solutions provide what Kirchhoff termed “cost governance” for intelligence about container expenses.
  • Serverless Computing: The serverless paradigm buttresses the multi-cloud movement by allowing organizations to “build services as necessary across different clouds with different approaches,” commented Stackery CTO Chase Douglas. Serverless methods fortify what Automation Anywhere Product Marketing VP Kashif Mahbub described as the cloud’s capital advantage of “the ease of use to get started.” Serverless computing lets organizations access, build, and manipulate web apps with cloud providers’ infrastructure. Leveraging JAMstack architecture with it “emphasizes speed and security,” added Ryan Coleman, Stackery VP of Engineering and Product.

Data Privacy, Security

The resounding worth—or risk—of any multi-cloud use case will always depend on preserving data privacy, which represents a modern metamorphosis of the security concerns that traditionally plagued this architecture. “The moment you mention cloud to somebody in banking, financing, insurance, the first question is: is it secure; is my data private?” Mahbub disclosed. In edge settings the issue is “how do I move data from an edge device to a hub in a secure manner,” DH2i CEO Don Boxley denoted; that cautionary point is equally applicable between clouds or between clouds and on-premises. The variety of multi-cloud milieus creates undue emphasis on what Boxley called the “different access policies” for governance and data privacy protocols.

Security methods for data in motion are exemplified by software defined perimeters, which synthesize public key access with encryption while “opening up a secure UDP port, temporarily, to form an [application level] connection, and then it goes away,” Boxley disclosed. Automation capabilities for container platforms resolve multi-cloud access “so you have an effective way how you can define, for different cloud providers or for different infrastructure or different types of cloud, policies in an automated way,” Scheele confirmed.

Private Clouds

The expanding adoption rates of private clouds is broadening and diversifying the multi-cloud ecosystem. This variety involves “companies running their own servers, their own environments,” Mahbub revealed. This trend is aligned with the multi-cloud trajectory because in some cases, organizations access cloud services “in their own private environment, and that private cloud could be on AWS or Azure,” Mahbub said. Cogent drivers for private clouds include:

  • Total Cost of Ownership: The operational expenses of public clouds can substantially accumulate over time, which is why “on a large scale it can be much cheaper to own your own infrastructure,” Kirchhoff acknowledged.
  • Regulations: The heightened regulatory environment makes private clouds attractive, since organizations obtain their collaborative and ubiquity benefits with unassailable control. “There’s more of an emerging thing, especially with privacy and regulations, where private cloud has a real role to play in the architecture,” Smith explained.
  • Control: Owning one’s own physical infrastructure gives organizations a degree of self-reliance that’s rare in the cloud.

Nonetheless, as Mahbub suggested, it’s oftentimes necessary to connect private clouds with other settings to fully benefit from this architecture.

The Edge

Multi-cloud’s enterprise utility (and mounting use cases) reflects the growing distribution of the data landscape. The varying degrees of isolation attributed to the present public health climate are redoubling this trend—and edge computing’s viability at the cloud’s fringe. Increasingly heterogeneous environments simply reinforce the need to connect them, which is where multi-cloud deployments excel.

About the Author

Jelani Harper is an editorial consultant servicing the information technology market. He specializes in data-driven applications focused on semantic technologies, data governance and analytics.

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: