Making Multi-Cloud and Hybrid Cloud Work Overview & Event Highlights

This event hosted by Brad Casemore, VP Research Datacenter & Multi-Cloud Networking at IDC, brought together industry “big guns” Vijoy Pandey, VP Engineering & CTO of Cisco Cloud and Oliver Cantor, Director, Verizon Global Products. They are joined by two highly successful CEO’s with amazing pedigrees which included the original SDN revolution, and now head-up two of the hottest Cloud start-ups of the moment; Steve Mullaney, President & CEO of Aviatrix and Amir Khan, President, CEO & co-founder of Alkira. This heavy weight panel debated the issues and challenges facing customers and vendors in the new Cloud era.

The following are some of the main takeaways from a lively, informative and at some points controversial debate. Links to a longer write-up and to a recording of the full Webcast and Podcast are at the end of this document.

 

Why multi-cloud?

  • Customers want to be able to take advantage of the different features and capabilities offered by different cloud providers – e.g. some applications run best in a particular environment
  • Customers want freedom of choice, increased business resilience and to avoid being locked-in
  • No single cloud provider offers complete, global coverage or all the services an enterprise needs

 

Customer needs and expectations

  • New players depend on cloud for agility, innovation and rapid growth. Established businesses are looking to cloud for efficiencies, increased velocity, new ways to deliver customer value and to arm themselves against competitive threats
  • Many businesses see digital transformation as essential to survival and cloud as the key to transformation
  • Enterprises are no longer prepared to wait for technology to catch up with their business requirements. Vendors are being driven not just the needs of their customers but by the demands of the customer’s customers
  • Enterprises need solutions that protect existing legacy investments and provide a single cloud-like experience across their entire IT estate, including existing on-prem environments
  • Businesses are no longer prepared to live with the trade-offs between the management control and SLAs that come with on-prem systems and the utility, usability and flexibility of cloud solutions
  • Enterprises need to be able to manage multicloud just as they would on-prem environments, with a ‘single pane of glass’ for end-to-end visibility across the entire network plus the ability to support monitoring, troubleshooting and other day two operations
  • Customers want to regain control of costs. The proliferation of spending on cloud applications and services has made it harder not easier to benefit from the apparent economic advantages of cloud, such as pay-for-use consumption

 

Challenges

  • Enterprises need to reduce the complexity of deploying and managing workloads and network resources in cloud and multicloud environments. Complexity reduces agility, increases risk, drives up costs and creates operational demands, eg on the skills of IT teams
  • Cloud environments are different and implement networking concepts in their own way. Inserting higher level services such as firewalls and active directory is not straightforward
  • Deep connectivity to a single cloud is challenging. Building and running multicloud networks is highly complex and time-consuming
  • Cloud-native applications require a new approach to development that has profound implications both for enterprise IT operations and for vendors
  • Lack of standards and limitations of existing technologies has slowed the deployment of critical cloud networking infrastructure
  • Cloud providers have exaggerated the capabilities of their services and done little to support customers to move to multicloud environments

 

Implications

  • Multi-cloud is an essential component of digital transformation strategies, but is more than an IT infrastructure. It is both a set of technologies and a new cloud-centric operating model
  • Migrating existing applications and workloads provides limited benefits. The new architecture demands new cloud-native applications and development methodologies
  • Truly distributed applications pose design and development problems but the pay-off is greater ease of maintenance, a smoother evolution path and a reduced ‘blast radius’ compared to today’s monolithic programs
  • Distributed applications running in multicloud environments require clean APIs up and down the protocol stack to improve performance and manageability, and reduce complexity
  • The concept of the traditional data centre is giving way to a distributed model – centers of data in the cloud
  • Some older and not-so-old technologies are becoming redundant including routers and SD-WAN
  • The role of the network in the cloud goes far beyond its traditional transport function, but encompasses all the infrastructure that enables control, visibility, security, advanced routing and high-level connectivity between applications

 

Want to find out more?

 

Article by Guy Matthews

Call for Papers

  • About You

  • About Your Paper Submission