WE ARE A SOCIETY INTENT ON GETTING more for less—two dinners for the price of one! Small children ride free! An extra large bottle of shampoo for the price of the regular size! What if you could squeeze four megabits per second through a T1/E1 WAN link? A bargain!

Examine the following graph. This company managed to double and triple the traffic through a branch office’s small WAN link. An appliance with acceleration features compressed the traffic, transferred it, and then restored it to its original state at the destination. Best of all, this company did not have to attend to any cumbersome administrative tasks such as defining and configuring compression tunnels.

Compression
Compression is a feature of application traffic management systems that increases enterprise network’s throughput, making it seem as if the WAN link offers more bandwidth than it really does. The best compression technology does three things:
1. Compresses traffic to squeeze more data through a limited WAN link without losing quality or data
2. Removes the administrative burden of having to define, configure, and maintain compression tunnels, the mechanism used to shrink, transfer, and restore traffic
3. Assesses real-time performance and latency data to adjust compression and bandwidth allocation to ensure optimal performance of critical applications

Just think of all the advantages to transferring more traffic through a limited WAN or Internet link. It enables enterprises to:
• Make critical applications faster
• Ease congestion on a saturated link
• Defer or avoid an expensive link upgrade or decrease existing link size
• Enable confident rollout of new enterprise-wide applications
• Survive when a provider is slow to deliver a link upgrade or when an upgrade is unavailable
• Reduce monthly charges if bills are based on number of bytes transferred

But be careful! It’s easy to overstate the benefits of compression, just like it’s easy to overstate the benefits of bandwidth upgrades. Critical applications that suffer poor performance aren’t necessarily the applications that get access to extra capacity. Usually, it’s the less urgent, bandwidth-hungry applications that monopolize bonus bandwidth. Combining shaping and acceleration techniques ensures that you get both more capacity and judicious use of your whole link.

Where Compression Devices Sit
Compression typically requires two appliances, one deployed at each end of a connection. Most organizations put one device at each branch office that exchanges traffic with other branches or a main site. This arrangement maximizes all bandwidth throughout the organization. Each device compresses its outbound traffic destined for a location with a compression partner. When the traffic arrives, the second device decompresses its inbound traffic and restores the traffic to its original state. These passageways between compression partners for compressed traffic are known as tunnels.

Compression Techniques
A leading compression technology should analyze passing traffic to see if and how it can make the traffic smaller. It should ask and answer the following types of questions and then take appropriate action.

Is this traffic headed to a destination where there is another acceleration device to restore the traffic to its original state? It should automatically detect its counterparts and create appropriate compression tunnels without requiring manual involvement.
What type of traffic is this and should we even try to compress it? It should automatically identify and classify each passing packet, independent of acceleration. It should then use that knowledge of each packet’s traffic type to determine if acceleration is appropriate. Previously compressed traffic (such as images, zipped attachments, and streaming media) and encrypted data (such as HTTPS and SSH) should not be compressed further. Such attempts result in little or no reduction in bandwidth consumption and increase latency and jitter.
Which compression algorithm would render the best results for this particular type of traffic? Frequently, applications benefit more from one compression method than from another. It should automatically apply the compression algorithm to each application that yields the most effective compression with minimal latency.

Typical Compression Results
A compression ratio portrays results in a standardized format. If 100 bytes of data is compressed into only 50 bytes, and it takes only half the bandwidth it would have needed, then its compression ratio is 2:1.
Some vendors promise 5:1 or even 10:1 compression ratios, but these figures are based on best-case tests. These types of statistics do not represent realistic expectations or results. A more realistic range for you to assume while planning or forming expectations is 2:1 to 3:1.
Not only do different types of traffic use different compression algorithms, they also get different typical results. One application might typically get a 4:1 compression ratio, while another might only get 1.5:1.

Automatic Tunnel Creation and Maintenance
Traditionally, compression solutions require you to manually define and configure compression tunnels. Let’s think about what that means. Suppose you have two offices and want to implement compression. You would only need to set up and configure one tunnel. No problem. Now suppose you have 20 offices in a meshed topology and want to use compression between each office. You’d have 190 tunnels to configure. Ouch!

Look for an automatic tunnel creation and maintenance feature that automatically detects acceleration devices on the network and builds acceleration tunnels between them. In addition, when resources are constrained and tunnel demand exceeds resources, it should automatically determine which tunnels yield the best results and ensure they remain active, while deactivating less utilized or less effective tunnels.

Latency Management
Standard compression adds latency and variability. Whenever shaping or bandwidth-management solutions and compression solutions are deployed separately in the same environment, this variability makes the whole traffic-management picture imprecise. Integrating compression and traffic shaping technology delivers both more capacity and the power to use that capacity effectively. The combination ensures that all precious network resources, including the expanded capacity, go to the applications that you consider urgent and important.

While traffic shaping prioritizes mission-critical applications and smoothes bursty traffic, compression enhances performance by fostering greater throughput, faster performance, and increased network capacity.