Once, there were only three types of clouds: public, private, and the nebulous hybrid cloud which, depending on whom one asked, entailed a synthesis of the first two or one of them enjoined with on-premise resources.
A number of more recent developments, however, have occurred to multiply the types of clouds and their purposes into the specialization befitting cloud computing’s preeminence in the contemporary data sphere.
The past two years have proved the cloud’s the essence of remote collaboration, for which it furnished a distributed workspace resiliency for organizations of all types at the very moment these capabilities were necessary to remain operational.
It’s also the de facto means of efficiently implementing most aspects of data-centric processing, from data management staples (as the proliferation of cloud data warehouses and governance solutions demonstrate) to the most advanced analytics (with numerous data science platforms deployed in the cloud).
The undue reliance on the cloud for managing data initially manifested as the hybrid and multi-cloud phenomenon. That movement is continuing apace to reflect a growing number of cloud types, including:
- Industry Clouds: Specific clouds with workflows, functionality, and regulatory considerations slanted towards a particular vertical have been pegged by Forrester as a critical development for 2022.
- Distributed Clouds: According to Gartner, the inter-cloud connections of distributed clouds enable organizations to master some of the traditional caveats of this paradigm, including regulatory compliance, data sovereignty, and security by locally accessing remote clouds.
- Regional Clouds: This cloud variety directly supports the cloud’s global presence by allowing organizations to manage their resources—like metadata—according to geographic regions.
These trends are pivotal to the enterprise in two ways. Firstly, they magnify the need to span clouds at will with cloud native constructs while adhering to governance, security, and compliance demands. Secondly, they illustrate the true transformation the cloud has wrought: enterprise infrastructure has become Software-as-a-Service.
“The cloud mindset is highly distributed, very programmatic; it’s software,” reflected Jacob Smith, Equinix VP of Bare Metal Strategy & Marketing. “It’s really infrastructure under the hood, but it’s software on top.”
From Multi-Clouds to Poly-Clouds
Perhaps the ultimate boon of the software infrastructure the cloud delivers is the ability to seamlessly transfer between clouds—for pricing, business cases, and specific applications—as needed. A burgeoning number of vendors (including Databricks, Snowflake, MongoDB, and more) are countenancing this need by rendering their services available in any public cloud for what Vendia CEO Tim Wagner termed “poly cloud goodness.”
Whereas multi-cloud was an insurance policy to avoid vendor lock-in, poly cloud is “a positive that makes it easier to collaborate and connect with partners who aren’t running on the same IT stack you chose,” Wagner pointed out. “Vendors have taken on the problems of cross-cloud connectivity themselves so customers don’t have to think about it.” This capacity expands the cloud’s cardinal value proposition of collaborative functionality while illustrating the commitment of major vendors—not just organizations—to fluidly transition between clouds.
Cloud Native Underpinnings
Although the complexity Wagner mentioned is masked by vendors for poly cloud deployments, it typically involves cloud native approaches like containers and Kubernetes. These methods (including the Infrastructure as Code tenet that underlies the software for modern enterprise infrastructure) have been endorsed by the major public clouds to further the multi-cloud phenomenon—specifically via Kubernetes. “Kubernetes is a powerful orchestrator that helps you move container workloads, but that’s not the whole story,” Smith explained. “It’s the brains under the hood that you can build an experience on top of.”
As such, the fact that AWS, Google Cloud, and Microsoft have devised their own open source Kubernetes distributions, enabling organizations to have an AWS experience, for example, regardless of where their applications are deployed—in other public clouds, industry, or distributed ones—is core to furthering adoption rates for spanning clouds. “Amazon has a Kubernetes distribution called EKS Anywhere that comes with a variety of security management tools,” Smith noted. “You can see it from your AWS control panel, even in your private infrastructure or another cloud. It provides visibility and management tools and the orchestration underneath.” Thus, no matter which cloud specialization organizations choose at the moment, they get the same trusted experience.
The cloud’s collaborative capabilities are useless without mainstays of data governance and security that typify traditional on-premise deployments. The capacity to swiftly shift settings requires even more stringent governance enforcement than there was before. In several ways, however, the cloud meets this challenge by serving as a centralized layer for IT to monitor the activities of distributed users for working at home, for example. According to Automation Anywhere VP of Global Market Catherine Calarco, “Because of the cloud you can deploy [applications] across the organization and use the same processes and rules from a central governing body that determines the guidelines.”
In this instance, the cloud betters centralized governance methods by serving as the means for valued oversight into remote, decentralized data use. Consequently, IT or governance personnel “can see what’s going on and service your business units no matter where they are,” Calarco observed. “This is critical for the expansion of the digital workforce and working at home.” There are even centralized governance platforms that propel access controls like obfuscation and others into distributed sources to fortify secure governance.
The mounting decentralization of the data landscape engendered by specialized and poly-cloud computing emphasizes the exchange of communication, data, and attendant resources for remote collaborations. These needs are ideal for multi-party applications fortified by distributed ledger technologies, which will quietly broaden their scope in 2022. “Organizations will buy something that gives them a business outcome like a data integration platform, which so happens to include distributed ledger technology,” Wagner predicted. “This ties into the SaaS delivery model and gives the opportunity for emerging technologies to make the business successful.”
Heavyweights like AWS and Oracle have recently introduced solutions equipped with blockchain as cloud offerings for their propensity—particularly when cloud native methods like serverless computing are involved—to reduce the latency of transaction validations. The result is use cases like a “high performance network to do 50,000 to 60,000 transactions per second,” Smith mentioned. Such throughput for credit card transactions, for example, makes more common deployments like supply chain management or delivery logistics seem mundane.
Distributed Cloud at the Edge
Distributed clouds exemplify inter-cloud connectivity. For example, distributed clouds allow firms to locally access a regional cloud through a public cloud provider’s Kubernetes distribution to get a safe, governed experience tailored to the needs of a parochial client. For an international company, for instance, “Human Resources might have core functions that need to be localized for different regions based on local government,” Calarco posited. In these and other use cases, the ability to access remote clouds locally yields several advantages. Smith commented that not only does the distributed model make the cloud cheaper, but there’s also “GDPR, compliance, and High Availability, and you’d have to destroy and have problems with like, 90 different devices in order to lose your data. So, there’s a lot of resiliency.”
This architecture also has substantial ramifications for edge computing because it decreases the latency of transmitting the results of edge processing. The aforementioned transmissions, similar to any of those at the edge, can be protected by segmenting them to prevent network takeovers by “not really letting them talk openly over the internet,” stipulated PDI Security Solutions Chief Security Officer Mark Carl. “If by chance there’s an attacker that gets into the IoT environment, they don’t have the path to launch some type of sophisticated attack.
Protecting the myriad cloud types organizations use is a foremost concern for cross-cloud computing. This issue is aggravated by endpoint devices at the network’s fringe in distributed clouds, the Internet of Things, and the edge. Digital agents are vital for these and other deployments. Gateway devices let firms “have a software agent onsite that can scan IoT devices and see if there’s anything that’s vulnerable,” Carl indicated.
Gateway tools also deliver file integrity monitoring, vulnerability management, and hash matching for remote patches. Moreover, bots buttress security in other cloud types by “anonymizing data,” Calarco revealed. “They can monitor and send automated communication for you with instant alerts to automate workflows so the right people know about security events at the right time.”
The cloud’s normalization—evinced most pointedly in its transformation of enterprise architecture to SaaS—across verticals has resulted in poly cloud computing, an expansion of cloud types, and a world of possibilities, like Artificial Intelligence programmable infrastructure. The most immediate gain, however, is the unprecedented ease of traversing clouds for business use cases.
“Workflows, especially post COVID, are getting harder and harder to operate in IT by pretending they’re an island,” Wagner concluded. “Amazon and others have said 80 percent of the business data they care about lives outside their four walls. You just can’t build an application in isolation any longer.”
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.