Comtech EF Data / Stampede
FX Series Administration Guide - Version 6.2.2
27
Chapter: Overview - FX Series
Section: Single-Sided Solution
MN-FXSERIESADM6 Rev 6
The drain algorithms are strict priority or Min-Max. In Strict Priority, available bandwidth is
allocated on the basis of priority.
Min-Max gives more control. Bandwidth is allocated up to a committed information rate based
upon priority. Once the committed information rate is reached for all classes, excess bandwidth
is allocated based on the same priority, up to a defined maximum for each Queue.
1.3.5
GZIP Compression
The most common use of compression in Web environments is accomplished by enabling GZIP
functionality at the Web server. GZIP compression is handled on-the-fly from the servers to the clients.
This reduces bandwidth consumption and improves application delivery and client response time. The FX
Series uses GZIP compression to reduce the payload size to deliver more data across the satellite link,
enabling more applications to be delivered and the ability to support more users. GZIP compression
removes non-essential information from data being moved from one location to another, and then
reassembles the data to its original form after the transfer is complete.
Squeezing the data reduces network traffic and accelerates the delivery of time-sensitive information.
GZIP compression uses standard techniques to compress data sent to browsers. While compression exists
in many forms throughout Web deployments, the FX Series is able to more effectively apply compression
resulting in better compression ratios. GZIP is not normally used for attachment compression or for
inbound compression from the browser. In addition, GZIP cannot be used to compress HTTP headers or
image data. In a single-sided mode, the FX Series appliance utilizes GZIP to compress information that can
be processed by standard browsers.
Stampede utilizes various compression techniques to reduce the amount of data that must be sent across
the network. In two-sided deployment, the FX Series bi-directional compression provides compression
for:
•
All HTTP Headers
•
Application Cookies
•
All Text and Data Objects
•
JPEG files with Image Reduction, yielding very acceptable quality
•
All attachments and file uploads and downloads
1.3.6
Image Reduction and Smoothing
Image Reduction and Smoothing reduces the amount of data required to represent an image without
significantly altering the visual perception of the image. This is accomplished in two ways. Smoothing
reduces the high frequency components or the sharpness of an image. A moderate amount of smoothing
can significantly reduce the amount of data. The quality factor of a JPEG image relates to the precision of
the samples. Sample precision can be reduced without visible detection.
The goal of the JPEG quality and smoothing values is to reduce the amount of data while maintaining a
usable image. Depending on the JPEG, the compression is often in the range 9:1. A number between 1
and 100 specifies the tradeoff between size of the jpeg data and quality of the original image. A higher
number will retain a higher quality but will not conserve as much bandwidth. If no value is specified then
the FX Series value is inherited from a higher level policy; a default value of 50 is used if no higher level
policy is defined. Images that have been transformed are typically not significantly changed by running
through the algorithm again. What this means is that if an image has been compressed with particular
smoothing and quality factor, if the same factors are used again, the image is not significantly changed.
