Today’s digital economy is an “as-a-service” world, where individuals and businesses have a plethora of choices in goods and services at their fingertips from multiple cloud service providers (CSPs). And in an increasingly multicloud world, this freedom of choice is being demonstrated by enterprises at a more granular level by selecting a CSP or service not just for each application, but by individual workloads up and down the application stack.
This select-sourcing of best-of-breed IaaS compute and storage resources across multiple clouds for individual workloads and applications is an accelerating and maturing trend in the enterprise multicloud journey. This new world requires faster and more agile inter-multicloud interconnection to keep up with the application requirements needed to transform and run today’s digital businesses.
For example, a finance department may break up a specific application workload by splitting its compute and storage requirements from different hyperscalers (e.g., compute instances from Microsoft Azure and storage from AWS) for the same application workload. Or it may target specific analytics applications to only analyze certain sets of data, running specific workloads on different cloud services, sourcing the best-of-breed compute or storage for each task. Different compute processes could run on multiple services from a single CSP, while data processing and analytics run in two totally different CSPs.
This privatized selection of public cloud services against specific workloads and processes requires fast, consistent and secure multicloud Interconnection to best leverage greater cloud performance, scalability (cloud velocity) and service chaining. In fact a recent market study published by Equinix, “The Global Interconnection Index,” projects that the installed Interconnection Bandwidth capacity required for global enterprises to privately interconnect to cloud and IT service providers is estimated to grow 160% by 2020 (see the diagram below)!
Harnessing inter-multicloud Interconnection
Inter-application connectivity among multiple cloud services is extremely difficult to accomplish over legacy, long-distance networks or the public internet. High-latency, unreliable, unpredictable and insecure connections cannot deliver the speed, consistency, performance or security required for highly integrated, process-to-process communications.
Here’s some of what’s needed to ensure continuous, integrated, inter-multicloud interconnection between or among multiple application workloads:
- A platform that can leverage access to multiple clouds at the edge of the enterprise, where most applications and data reside
- Private, direct, secure and predictable connectivity between and among public CSPs and cloud services
- Flexible, low-latency, high-bandwidth connectivity that can be dynamically spun up and down as needed by application workloads
- Proximity and adjacency of compute and storage where there is the highest concentration of people data and apps, so there’s no need to traverse long-distance networks for the greatest performance (aka, “the headquarters experience” is available to anyone in the enterprise)
By providing a more flexible and dynamic multicloud environment and inter-multicloud connectivity when splitting application workloads, you can leverage greater cloud velocity and service chaining scalability.
Creating the SuperCloud
Smithfield Foods has been in the process of building a “SuperCloud,” as described in a blog posted by its director of cloud and solutions architecture, Matthew Douglas. According to Douglas, the SuperCloud design “gives the global enterprise (user, group, truck, plant, headquarters) the ability to geo-route to the best performing cloud or most cost-effective cloud for each application.”
The primary reasons that Douglas opted for multicloud, cross-cloud scaling were:
- Fault tolerance for high availability applications
- Ease of maintenance
- Combined strength of each cloud service in a single service
Smithfield Foods’ super cloud leverages three cloud services: Microsoft Azure, AWS and Virtustream. The Microsoft Azure, Office 365, Active Directory and Single Sign-On capabilities gave Smithfield Foods the ability to route and replicate traffic from region to alt-region without leaving the Azure network, enabling the company to pull together international divisions with minimal costs. AWS gave Smithfield Foods the largest number of SaaS solutions to choose from.
“The true techie inner geek just says you have to use it,” Douglas said.
Virtustream’s ERP (SAP) hosting with a native VMware format gave the company easy scalability of on-premises VMware resources on the Virtustream cloud, and the ability to put legacy systems on the same network as cloud resources.
The inter-multicloud Interconnection was enabled by Equinix and its Cloud Exchange, which provided economical, direct and secure cross connection to all the clouds and Smithfield Foods’ MPLS network within a single Equinix site. “Truly the only choice,” according to Douglas. You can check out an overview the Smithfield Foods’ SuperCloud by going to Douglas’ blog.
Whether you are placing various applications in multiple clouds, or spreading a single application’s many workloads across them, leveraging direct, consistent and secure interconnection will enable your company to create its own super cloud for increasing your application performance and scalability.
Learn more about harnessing Interconnection for your multicloud environment by downloading the white paper, “Home of the Interconnected Cloud.”
Also, hear more from Matt Douglas at Smithfield Foods and Aon, Devon Energy and Lincoln Financial and about how they leveraged Interconnection to build their digital businesses.