Jan 19, 2012
One of the most talked-about new buzzwords in the past year is “big data.” A recent Forbes article points out that big data is not just quantity, but also includes multiple types of data. Having a lot of data sitting around doesn’t really accomplish anything; the real key to big data is being able to analyze large diverse data sets and act on the results. While WAN optimization can’t help you analyze the data, it can help you move the data to the right place as quickly as possible and with the lowest bandwidth cost.
Most of the focus on big data from storage companies concerns how to store, protect and guarantee availability. Analyzing the data is quite a bit more difficult, usually requiring clusters of servers. With large clusters commonly used in enterprise deployments, on-demand cloud computing is typically mentioned in the same breath as big data.
However, there is one problem with analyzing big data in the cloud: moving the data to the cloud.
Moving big data into the cloud means crossing two big hurdles: location and bandwidth. First, the farther away the cloud data center is from your site, the more latency you have to deal with and the longer it will take for your data transfer. Second, bandwidth is important as well, since insufficient bandwidth means the data transfer will take an excessive amount of time. Add the transfer time to the analysis time and it is possible that the resulting big data analysis will be stale and outdated by the time everything is finished.
As Edd Dumbill from O’Reilly mentioned in a recent article, this brings us to the third hurdle in moving big data to the cloud: velocity. With an exponentially increasing amount of data coming into an organization and skyrocketing analysis requirements (think “streaming data”), incoming data must be analyzed as quickly as possible.
If you are using cloud computing to analyze your big data, and you happen to be located in the same city as the cloud data center, and you have unlimited bandwidth, you are ready to go (and probably aren’t reading this post).
If, however, like most people, you are dealing with latency and limited bandwidth, WAN optimization can help. All of the features that help business move and access data across a WAN also apply to big data movement into the cloud.
Replicating big data over a WAN has the same problems as moving data into a cloud data center. Meeting any replication requirement can be difficult and, as the size of the data grows, so does the complexity. Silver Peak has a long history of optimizing replication over WAN connections, with virtual appliances that scale to 1 gigabit-per-second (Gbps) in WAN capacity and physical appliances that scale to multi-Gbps. Replicating big data is no different. If big data replication is a requirement, Silver Peak can help.
Whether you are trying to move big data into the cloud or replicate it across a WAN, Silver Peak has a solution that can help. Network Acceleration can overcome the impact of latency, Network Integrity can correct lost and out of order packets, and Network Memory can reduce the amount of bandwidth required to move big data. Big data CAN be moved to the cloud — and Silver Peak can make it a reality.