Peribit drops unnecessary WAN traffic

We mentioned last time that quite a bit of activity has been going on in the bandwidth optimization space. One of the more interesting announcements made by Peribit Networks recently was a capability it calls Network Sequence Mirroring, which augments the company’s traditional Molecular Sequence Reduction compression technology.

We mentioned last time that quite a bit of activity has been going on in the bandwidth optimization space. One of the more interesting announcements made by Peribit Networks recently was a capability it calls Network Sequence Mirroring, which augments the company’s traditional Molecular Sequence Reduction compression technology.

Network Sequence Mirroring combines compression and caching to reduce the volume of traffic you send over the WAN to boost application performance and reduce WAN bandwidth usage and costs.

Peribit’s SM-500 device, due to ship in August, “stores hundreds of gigabytes of sequences and is good [at accelerating] applications like e-mail and [network] storage,” says Mike Banic, vice president of corporate marketing at Peribit.

For example, he says, a PowerPoint presentation sent as an e-mail attachment can get cached. If subsequent versions of that PowerPoint file are resent multiple times and sequences in those files are compared to the original – and just a page or a bulleted item on the page has changed – the stored sequence that actually changed gets flagged. Then, only the changed sequence in the attachment need traverse the WAN, Banic explains.

He adds that users can selectively enable the mirroring on an application-by-application basis.

The SM-500 costs between $9,000 and $40,000, depending on the size of the WAN link connected to it. Connections supported range from 256K to 20M bit/sec.

Peribit also announced Version 5.0 of its Sequence Reduction System (SRS) that supports policy-based multi-path capabilities. This means Peribit compression/caching appliances can dynamically monitor dual-homed WAN connections to a remote site. Users can set policies that describe the latency and packet loss requirements for a particular application. The SRS appliances can detect which link cuts the mustard for each application at a given moment and direct the router to forward the application traffic over it.