There are two big pieces to implementing cloud computing services. The first is a robust data center that is sized to hold all of the equipment anticipated, and supported by an ecosystem of cooling, backup power and security. The second is connectivity. If you don’t have a reliable way to deliver your cloud services, it doesn’t matter how groundbreaking they are. Enterprise customers need to know that they can count of the system being available and functioning at all times without slowdowns or dropouts.
This gives the network bandwidth providers a leg up in making their case for high performance cloud solutions. A good example is EarthLink Business. They were in the metro and long haul network business long before clouds started to appear. EarthLink operates a high speed nationwide network that spans 28,00 fiber route miles with 90 metro fiber rings. This network supports six classes of service for MPLS over Ethernet, T-1 and DSL connections. Class of Service (CoS) is honored from edge to edge over the entire network.
Why is that important? Today’s converged networks are only effective if they can handle data, voice and video seamlessly. Voice over network communications, like VoIP telephony, is especially sensitive to latency, jitter and dropped packets. You need CoS to provide a fast lane for real-time two-way services like VoIP so they aren’t destroyed by mundane file transfers and backups hogging all the bandwidth.
This argues for a high performance privately operated network over taking your chances on the public Internet. A private network operator can carefully allocate resources to ensure that every packet stream is supported to ensure high quality operations. On the Internet, there are no such guarantees. You launch your packets and you take your chances. For one way video, email and Web browsing this is perfectly adequate most of the time. It tends to fall short for enterprise VoIP, video conferencing and cloud access.
Why is connectivity with the cloud so fussy? If all you are doing is backing up files in a background process, you may never notice any issues. However, if you have moved from your own data center to the cloud and are your applications are now running as SaaS (Software as a Service), any delays in response through the network will be frustrating at least and a detriment to productivity at worst. You moved to the cloud to save money. It’s a shame if poor network performance eats up those cost savings with reduced employee productivity.
EarthLink Business and other fiber optic network owners see the competitive advantage they have by providing both the cloud resources and connectivity as a single vendor. That’s why you’re seeing them expand into the cloud space as fast as they can. EarthLink is building four more data centers to support its next generation cloud hosting platform. These will be located in San Jose, Chicago, Dallas and South Florida. An existing data center in Rochester, NY will be updated early next year.
The fiber network is also being expanded to accommodate higher traffic levels anticipated by cloud users. These include key East Coast markets such as Ashburn, Atlanta, Charlotte and Orlando plus new capacity for the Texas cities of Austin, Dallas, San Antonio and Houston. The entire network is being beefed up to support the new native benchmark of 100 Gbps. Businesses today can typically get fiber optic service up to 10 Gbps in major metro areas.
Do you see the advantages of moving your IT operations to the cloud, but worry that you’ll lose performance in the process? Take a closer look at enterprise grade cloud computing and networking services and see how they have raised the bar on cloud performance.