Product and service reviews are conducted independently by our editorial team, but we sometimes make money when you click on links. Learn more.

Public Clouds: Matters Of The External Connection

Public Clouds: Matters Of The External Connection

Public Clouds: Matters Of The External ConnectionPublic Clouds: Matters Of The External ConnectionWhen migrating to a public cloud, few IT departments would dispute that security remains as much of a concern as it was when applications were hosted on-premise. Similarly, few would also argue that network bandwidth continues to be critical, especially external bandwidth between the organization and its chosen public cloud service(s). As John Sloan, Info-Tech research analyst, says, it’s a mistake for enterprises to assume that after making the move to a public cloud service that all “infrastructure has gone out the door” and bandwidth worries are no more.

Although enterprises’ concerns about public clouds today are arguably less focused on bandwidth than on availability/reliability, this doesn’t mean that enterprises are off the hook where the network is concerned. The reality, Sloan says, is that “cloud vendor service guarantees focus on the availability of the service, but a ‘five nines’ available service is of little use if you have a ‘three nines’ network connection.”

Here, the critical link is the last mile from Internet to desktop.

Because redundancy translates to greater availability, enterprises should carefully consider the level of redundancy required in its network to ensure service level targets are met. This extends to component redundancy (switches, etc.) but also carrier redundancy, including knowing what uptime guarantees the carrier handling your Internet connection provides. To guarantee high availability, redundant carrier connections might be necessary—something that will add to the total costs of using a public cloud service but adjusts for outages in areas that cloud provider has no control over.

Typically, says Michael Shonholz, CDW telecom services manager, deficiencies in an enterprise’s external connection to a public cloud reside at remote sites, with the majority of outages due to last mile issues with the local exchange carrier. “Hosting all the applications in a client facility provides the potential for uncontrollable deficiencies in the network,” he says. Using an enterprise, carrier-neutral facility for all computing, however, means tapping into a facility built on fiber Internet and that typically has built-in redundancy. Such settings typically feature fewer last-mile connections.

Just when an enterprise will deem redundant connections necessary depends on the applications being put in the cloud and their tolerance for uptime. Two connections (primary and backup) should be sufficient, assuming true carrier diversity is possible, and the application will determine the bandwidth required.

Meanwhile, the level of security for these connections also generally depends on the application and who is providing the connection. Many carriers provide both a router and device management. When weighing a dedicated connection against a public Internet line, note that the public Internet is ultimately a best effort proposition that comes with no guarantee of speed or ability to provide quality or class of service. The pros, cons, and limitations of each connection type are dependent on the cloud service being purchased and the tolerance for outages, downtime, or accessibility.

Also note that there are connection options available beyond those from leading carriers, including pure IP suppliers that provide unique bandwidth availability to enable cloud usage. “Clients will define their tolerance for outages and availability,” Shonholz says. “In today’s environment, you can design a very cost-effective network that provides high availability and uptime. The sky is the limit in terms of cost, but right-sizing the WAN based on application use and client tolerances will dictate the final requirements.”

William Van WinkleWilliam Van Winkle

William Van Winkle has been a full-time tech writer and author since 1998. He specializes in a wide range of coverage areas, including unified communications, virtualization, Cloud Computing, storage solutions and more. William lives in Hillsboro, Oregon with his wife and 2.4 kids, and—when not scrambling to meet article deadlines—he enjoys reading, travel, and writing fiction.

See here for all of William's Tom's IT Pro articles.

Check Out These IT Videos