Page 1 of 2 next >>

Enterprises Learn to Unravel the Complexities of Multi-Cloud

Multi-cloud is changing the way we manage data. It’s seen as a way to build a more resilient diversity of services while ensuring a greater degree of independence from a single vendor. At the same time, it takes skill to get everything aligned. In recent years, multi-cloud has become a popular approach, with 93% of enterprises using a multi-cloud strategy, according to the latest Flexera/RightScale survey on cloud adoption. The survey found that respondents use an average of 2.2 public and 2.2 private clouds.

It’s important for data managers and their business counterparts to understand the benefits and implications of the cloud implementations they are using to run their businesses. “Enterprises don’t choose multi-cloud,” said Sriramkumar Kumaresan, head of service lines markets at Mindtree. “Multi-cloud chooses an enterprise as a result of keeping up with dynamic and multiple dimensional business transformations. An example is mergers and acquisitions. The advantage of a multi-cloud environment is ensuring availability of the right applications at the right place, at the right time, with the right speed and at the right cost.”


What is driving the appeal of multi-cloud arrangements? Avoiding vendor lock-in is a frequently cited advantage to multi-cloud strategies, said Matt Quinn, chief operating officer of TIBCO. “It allows for different architectural and technical benefits that wouldn’t be available if the organization uses a single cloud vendor. Multi-cloud arrangements allow organizations to invest in different vendors for different purposes, encourage architects to think about core use cases and make sure services are portable, and open up the possibility for new services that a cloud of choice may introduce in the future.”

Embracing a multi-cloud strategy “enables true freedom and control to run an application, workload, or data on any cloud,” agreed Paul Speciale, chief product officer at Scality. Modern enterprises have sophisticated needs that can no longer be satisfied by one cloud provider, he added. A multi-cloud strategy ensures a higher level of data availability and durability because data can be replicated across multiple, fully autonomous clouds. Another benefit is cost savings, due to greater competition between cloud storage pro-viders and the ability to instantly take advantage of new pricing for any given storage offering, added Speciale.

Increasingly, “the leading driver for a multi-cloud strategy is maintaining flexibility and choice of providers; enterprises want to avoid vendor lock-in and take advantage of best-of-breed solutions,” said Vikas Mathur, senior vice president of products at Actian.

What does it mean to be multi-cloud? “Some think it is running an application across multiple cloud providers so that they can capture cost benefits or reduce risks,” said Jim Walker, vice president of product marketing at Cockroach Labs. “It is pretty difficult to pull this off though, as technical complexities are pretty intense. Hybrid cloud for many is a reality and can be considered multi-cloud as well. Often in highly regulated industries or with high-value workloads, organizations will want better control.”

There are availability concerns that make a multi-cloud strategy more preferable as well. “Not every cloud vendor offers services in every country or region, and having multiple cloud vendors gives customers more geographic coverage,” said Quinn.  Another way to think about multi-cloud is multiple regions, which can be complex but a very pragmatic reality for many, as they want failover or they need applications to live in different regions to meet latency requirements of their users, Walker said.

In the era of digital transformation, data is playing a more important role than ever, said Rick Vanover, senior director of product strategy with Veeam. “A multi-cloud approach enables companies to take advantage of many benefits, including increased agility and cost management,” Vanover said. This also aligns to developer skillsets, since many organizations that have multiple lines of business or divisions will inherently have a multi-cloud strategy based on their different areas of expertise.


Applications that are typically deployed across multiple cloud environments tend to be data-intensive, said George Burns III, senior consultant for cloud operations at SPR. “More common multi-cloud implementations are with data science, business intelligence, and data warehousing engagements,” he said. “As high-growth sectors in recent years, these data-dependent areas have flourished by treating data location as an agnostic attribute, creating connections between platforms to use data in place without the need to migrate or copy data from its original location.”

According to David Tareen, director of AI and analytics at SAS, multi-cloud is a natural environment for analytics “due to the bursting and unpredictable nature of analytic workloads. Analytic applications are often deployed across multiple cloud environments to take advantage of scale. To eliminate data movement, we often see analytic applications being co-located with the data across multiple cloud environments.”

Multi-cloud also serves as a supportive environment for “application ecosystems that comprise multiple applications and/or tools,” said Mike Rulf, CTO of the Americas at Syntax. “For example, due to a geographically dispersed workforce, an organization may turn to virtual desktops as a way to ensure secure access to the corporate environment.” Once on that virtual desktop, a VPN or private connectivity can be used to access a corporate ERP system that could be on-premise or hosted with a private cloud provider, he explained. “In turn, that ERP system may tie to a SaaS solution for HR data and integrate with a data lake for dashboards and visualizations.”

Choosing multi-cloud application workloads requires understanding which deliver strategic and dynamic growth and prioritizing those for cloud deployment in a multi-cloud strategy, said Janine Corey, global head of cloud solution architecture for Qumulo. “The most critical step is to have a clear business value-based strategy and cross-functional leadership alignment around various workloads’ contribution to the overall business,” Corey said. The idea is to move to the cloud those workloads that benefit from scaling up and down dynamically. “Applications that drive compelling digital customer experiences and are core to a company’s brand, growth, and expansion strategy should be not only moved to the cloud but also modernized to take advantage of fast scale-up for unexpected big growth spurts, and the ability to be elastic for the best cost efficiency, offered by the cloud.”

Page 1 of 2 next >>


Subscribe to Big Data Quarterly E-Edition