Optimizing WAN throughput is a major focus of government and corporate IT departments. Latency, packet loss and bandwidth limitations (to name a few) have traditionally been determining factors for how much and how fast data can be moved down a pipe. That is no longer the case. Accelerating data transfers for long distance WANs is now a reality.
George Crump explains in a recent article: “The four walls of the data center no longer confine users, applications, and data.” – Organizations are multi-site, and users want to work from anywhere. Making ‘work from anywhere’ a reality still requires data and data still has gravity. Data has to move to the organization’s various offices and into the cloud. Data movement requires WAN bandwidth, and while bandwidth speeds have increased, the speed of light has not, latency as well as packet loss, are still critical factors in WAN performance.
Deduplication is not WAN Optimization – Most WAN optimization solutions leverage deduplication and compression to improve WAN performance. The problem with counting on data efficiency technologies to improve WAN performance is that they do not improve WAN performance; they improve data efficiency. Moving less data is not the same as moving data more efficiently, and of course, data efficiency technologies require redundant data to work. Data efficient WAN solutions do little if anything to optimize the on-the-wire performance of data transfers.
Fixing the WAN
WAN connections, as they get further from the source, introduce latency. It doesn’t matter how much
bandwidth is available, as distance increases so does the chances for packet loss, which requires a
retransmission. The key to getting the most out of the WAN is to work around both of these issues.
The first step is to increase the parallelism, which means sending more data over the network at the
same time from different sources. An efficient WAN optimization solution will increase parallelism even
from a single workload. The second step is to set transfer packets size to reducing the frequency of
packet retransmission. The problem with optimizing packets is that the “right” packet size constantly
changes on a WAN.
Introducing Bridgeworks – True WAN Optimization
Bridgeworks’ core technology enables organizations to achieve almost maximum bandwidth utilization
despite any potential latency. They accomplish this by creating a high level of parallelism and by
optimizing packet size for the given state of the WAN. The solutions use an artificial intelligence engine
to analyze the condition of the WAN connection continuously to determine the optimal packet size
along with the level of parallelism so the minimum number of packets needs retransmission but at the
same time maximizing the data throughput.
The Artificial Intelligence is self-learning, managing and configuring. IT doesn’t have to make any
changes to existing applications or workflows nor do they have to install agents. The Bridgeworks
solutions do not touch any data, and they optimize any data type; deduplicated, encrypted or
Bridgeworks solves the WAN problem so differently it may be time for a new category, instead of
WAN optimization, Bridgeworks’ delivers the solution for accelerating data transfers for long distance WANs in real time.