WAN Optimization Critical to Virtual Data Environments
Software defined networking appears poised to make probably the most radical change in the enterprise to date. In the very near future, organizations will be able to spin up entire data environments on fully virtualized infrastructure, which will most likely be distributed across great distances.
This will turn some very basic assumptions about data networking on their heads. Most notable? The breakdown of the data center edge and, in a larger sense, the LAN and the WAN.
For the most part, enterprise utilization of the cloud is limited to storage operations. But increased reliance on SaaS, PaaS and IaaS architectures will almost certainly drive a push to deploy higher-order applications, perhaps even mission-critical ones like ecommerce and financial reporting, over broadly distributed architectures.
At that point, enterprises will encounter a number of problems. In the first place, wide area networks simply are not designed for the highly dynamic data traffic that populates the local area network. Traditionally, the WAN shunted large blocks of data to field offices and remote locations where local infrastructure then took responsibility for parceling it out to select locations. Secondly, wide area services are not cheap. Even though worldwide bandwidth should suffice for the foreseeable future, the average enterprise may still balk at the cost of rising WAN traffic, even with robust optimization technologies in place.
But in true capitalistic fashion, where there is a need, someone will find a way to make a buck. And the need to improve WAN performance for cloud-level application environments has led to a bumper crop of startups peddling a wide range of innovative solutions.
Mountain View, CA-based cloud provider Axcient, for example, recently released a new sync protocol that aims to speed up data transfer for its Recovery-as-a-Service offerings. The system features a local appliance that keeps track of block-level changes within the enterprise environment, reducing data volumes to the cloud some 90 percent and, in turn. cutting the time needed for backup and recovery processes from hours to minutes. Axcient says it will integrate the protocol into its general cloud services portfolio next month, where it will be available at no extra charge to existing customers.
Meanwhile, a company called Pertino is looking to simplify WAN operations by adding more functionality to the network edge. The Network Service Virtualization (NFS) system, which the firm bills as a service-based approach to Network File Virtualization (NFV), converts traditional edge functions like mobile VPN, malware detection, monitoring and even WAN optimization itself into cloud-based services. This not only reduces complexity in the local infrastructure, but increases application and user functionality by decentralizing the network functions library, offering unified configuration, data collection, policy enforcement and other tools across multiple data environments and on multiple devices. Look for NSV as a key component in the newest version of Pertino’s Cloud Network Engine.
And from CloudOpt comes the new Speed-as-a-Service offering, which improves data transfer speeds to and from popular cloudservices like AWS and EC2 by 90 percent. This allows for near real-time replication of databases across regions, as well as full daily backup and high-speed upload of Big Data volumes to available cloud resources. What’s more, the system, accessed through an easily installed Windows or Linux client, requires no hardware or software appliances, nor any networking experience to operate. CloudOpt recently expanded the service to the U.S. west coast and the European Union region.
The rise of WAN optimization services is a key example of how the cloud can be leveraged to solve some of the very problems it creates. WAN functionality, already a critical component of the enterprise data environment, will play an increasingly vital role as organizations shed the bonds of traditional data center infrastructure. And that means it will have to become increasingly LAN-like in its performance capabilities.
The data world is getting smaller, and as the enterprise becomes more steeped in the cloud, all infrastructure will soon be local.