Jun 23, 2009
Cloud computing is a compelling way to deliver web-based and non-web-based applications that better utilize the physical infrastructure, while lowering costs by moving from silos of expensive customized servers to racks of cost-effective ones. Due to the elastic nature of clouds, a particular application may be delivered from one physical location at a given moment but from a different, potentially far-away location in the same cloud at a later moment. Performance and scaling of cloud computing is very important for all modern web applications due to the high bandwidth needed on a per transaction basis, number of transactions and transaction rate.
What implications do new interaction models have on L4-L7 application delivery services infrastructure? For one, a high degree of scalability and handling of peak loads will be required as larger sets of users access cloud applications. Second, the always-on nature of cloud applications requires resiliency in L4-L7 cloud networking services. Third, the agile nature of cloud applications themselves require that cloud-networking services be dynamically provisioned and �follow� application workloads wherever they meander. And finally, the fundamental cloud tenant of efficiency requires that cloud network services be shared across different applications and user communities without unduly increasing the physical infrastructure and/or its complexity.
A thoughtful and seamless L2-L7 cloud networking services foundation with breakthrough price-performance enables 1Gb/10Gb Ethernet to be deployed everywhere in the data center, which can significantly improve server utilization and consequently data center power efficiency without compromising application performance.