Multi-Cloud is a Must

By Matt Wallace, CTO, Faction

At the start of this year, few organizations fully recognized the power of being cloud-enabled: the ability to turn on a dime and successfully refocus the business in a new direction to meet rapidly changing needs. Yet this became the most significant business lesson of 2020's wild ride: It's not necessarily the strong who survive, but those who can adapt quickly.

In 2021, IT will continue enabling business' adaptability through agility and enabling access to the best innovations across clouds. The best way to deliver on this is through multi-cloud, which permits teams to take advantage of frenetic innovation across clouds, while minimizing or eliminating the cost to move data, and enabling the use of optimized hardware, including spot resources, across clouds.

A New Architecture

The demand for multi-cloud will be so strong in 2021 that more than 90% of enterprises (globally) will be using "a mix of on-premises/dedicated private clouds, multiple public clouds, and legacy platforms," as estimated in research from International Data Corporation (IDC). But multiple clouds is not multi-cloud, and the key goal of IT leaders should be to ensure that their organization has a multi-cloud strategy, not a multiple-cloud unplanned sprawl.

Implementing multi-cloud well requires a coherent architecture to minimize complexity, creation optionality, and perhaps most of all, optimize for high throughput and low latency to ensure that data gravity doesn't cause issues. As app teams select specific capabilities from various clouds to meet specific development goals, IT teams need an effective plan to handle access, cost, and data mobility concerns.

The right multi-cloud strategy supports these teams by providing new efficiencies and preventing unnecessary duplication of data (or its associated costs). Structured appropriately, multi-cloud can provide for the management of data from a central location, accessible to all clouds simultaneously, with high-speed data access over low latency connections. Multi-cloud then facilitates innovation by freeing users from the constraints (including services and data extraction or egress expenses) of any single cloud, while preventing data gravity issues, which can significantly slow and limit the ability to move or work with data. Strategies to integrate cloud, edge, and data center footprints will see a boost when the penalties for moving data are reduced.

Multi-Cloud Data Services

Organizations that moved to the cloud for agility are now finding themselves confined by individual clouds. Can you settle for the innovations of just one cloud if that only represents 40% of the innovation? Probably not. Your cloud strategy is an investment in agility and innovation; you can't afford to have only half of that loaf.

Every development team needs to be empowered to do what it does best with the tools that are most appropriate. Multi-cloud data services provide enterprises with a single, managed copy of data. The alternative-multiple copies of the same data scattered across clouds and on-prem- creates data synchronization issues and exponentially drives up storage costs. And avoiding dealing with multi-cloud results in shadow multi-cloud, just as shadow IT initially drove cloud adoption.

Drivers of Multi-Cloud

The flexibility provided by the wide array of cloud data services is one of the main reasons enterprises are moving to multi-cloud. Taking a closer look, multi-cloud will continue to be driven by advances in:

• Edge Computing: Data needs to be urgently available and centralized for some of the technologies that are growing most rapidly, with use cases for 5G and connected cars.

• Large-scale Analytics: Supercomputing performance moved well beyond the 300-petaflop mark in 2020, thanks in part to projects like the COVID-19 High Performance Computing Consortium. Projects with such a massive scale require multi-cloud. This necessity will grow in 2021, as enterprises leverage cloud and application performance advancements to rapidly gather insights from data. Faction recently completed an at-scale genomic data processing effort, demonstrating that there is potential for huge gains in performance and costs from a multi-cloud architecture for even the most intensive workloads.

• Advances in Data Warehousing: Look no further than the December 2020 AWS re:Invent announcements to see that the demand for data warehousing continues to grow. Faction showed this with Yellowbricks, where our combined solution showed amazing TPC-HS results. As the amount of data in data warehouses grows, the demand for that data to be leveraged across projects will grow-and without any guarantee of which cloud the new projects will leverage. Additionally, announcements like AWS Babelfish and Google's multi-cloud support in BigQuery show that cloud providers will continue to do their best to make data usable on their platform-wherever it is in whatever format.
  

We expect incredible technological advances in the coming years. Autonomous drone delivery will become more feasible, and even normalized as we wait for the COVID vaccine to be rolled out universally. Continued investment and research into autonomous driving brings us ever closer to a level 4+ self-driving car. More and more Americans will have robots in the home. But one foundation will help facilitate the development of them all: multi-cloud.

##

About the Author

Matthew Wallace is Chief Technology Officer at Faction. Matt has 25 years of service and product engineering in the tech field. His work includes frequent speaking, patents, and articles. He is the co-author of Securing the Virtual Environment: How to Defend the Enterprise Against Attack, one of the first books to holistically address cloud security concerns.