alphaspirit - Fotolia
The growth of cloud computing has introduced a series of advances to virtualization and on-demand resource allocation. Starting with Amazon's release of S3 object storage, cloud providers have enabled an array of IT services. And, although it gets less attention than compute, storage and higher-level software services, networking is no exception.
With network as a service (NaaS), customers access third-party network transport services over the Internet and pay for them in a subscription-based model. In most cases, the network provider uses protocols such as OpenFlow to implement software-defined networking. This allows providers to allocate resources based on users' changing needs.
Exploring three common NaaS offerings
A virtual private network (VPN) is one of the most widely used NaaS offerings. VPNs allow organizations to extend the function of their secure, on-premises networks to insecure networks such as the Internet. To protect the integrity and confidentiality of the data traversing those insecure networks, VPNs implement additional security measures. While similar to wide area networks, VPNs do not depend on dedicated communication links between networks.
Bandwidth on-demand is a service that allocates network resources based on the needs of particular devices on that network. More bandwidth is allocated to nodes experiencing peak demands, and, as demand subsides, bandwidth is scaled down.
Major network and telecommunication providers such as Verizon, CenturyLink and XO Communications provision bandwidth on-demand services, which complement compute and storage on-demand. Imagine, for example, a retailer that experiences peak traffic during the holiday season. Architects for the retailer would use techniques like autoscaling to meet the demand for compute resources. To ensure customers experience low latency, the architects would use elastic storage services or a database as a service offering, as well. However, if there is a bottleneck in network capacity, the benefit of this model is easily undone.
Cloud providers provision bandwidth at sufficient scale to make sure traffic into and out of the cloud is not a concern. A more vulnerable area -- especially for businesses deploying hybrid cloud -- is the connection between the cloud and on-premises environments. Cloud applications may need to access data or services running on-premises or in a private cloud. This connection can become a bottleneck and introduce unwanted latencies in application performance.
By adjusting the amount of data that can be transmitted over a network -- or essentially providing a bigger "pipe" for data to travel -- bandwidth on-demand helps address this issue. An alternate solution is to reduce the data volumes that have to be transmitted, which is where transport optimization comes in.
One way to minimize the amount of data transmitted over a network is to use a content distribution network (CDN). CDNs cache content across geographically distributed nodes. When users access static content from a service, they are routed to the closest available CDN node. If the requested content is cached on the node, it is delivered to the user immediately. Otherwise, the content is retrieved from the source system, delivered to the user and then cached for future reference.
There are many CDN providers, including Amazon, Comcast, Level 3, Microsoft Azure and Verizon. The caching approach works particularly well for frequently requested static content. It is less useful, however, for dynamically generated content.
Dynamically generated content benefits from protocols that are optimized to reduce latency. For example, a network provider may implement protocols configured to transmit large packets, or optimize error-checking protocols to minimize transmission overhead. This type of network optimization is especially useful when transferring data globally. Akamai and CDNetworks are two network providers with global reach that offer advanced services.
Businesses using the cloud for on-demand scalability need to consider network bandwidth along with compute and storage resources. Cloud providers like Amazon, Google and Microsoft are well suited to offer scalable services and storage, but third-party network providers offer critical cloud-based and on-premises resources to support a company's IT platform.
About the author:
Dan Sullivan holds a master of science degree and is an author, systems architect and consultant with more than 20 years of IT experience. He has had engagements in advanced analytics, systems architecture, database design, enterprise security and business intelligence. He has worked in a broad range of industries, including financial services, manufacturing, pharmaceuticals, software development, government, retail and education. Sullivan has written extensively about topics that range from data warehousing, cloud computing and advanced analytics to security management, collaboration and text mining.
Test your cloud applications faster with these tools.
Are networks a roadblock to progress?
Optimize network bandwidth for the cloud.
Is networking infrastructure the Achilles' heel of cloud computing?