About Axis Data: Market Landscape
As enterprises strive to speed content delivery and improve customer service to partners, employees and suppliers the emphasis on achieving faster online performance has never been as important. Tasked with delivering information in "Internet time" while providing users a 'live' Internet experience has forced companies to evaluate their web performance. IT departments are looking for new ways to accelerate web site performance in order to deliver faster, more responsive information to their users. Organizations are looking to achieve an end-to-end performance boost all the way from the web server in the enterprise data center through the "last mile" to the end user.
Most web sites today do not deliver fast speeds for users and the primary cause is insufficient bandwidth connections. The solution to the problem lies in technologies that optimize bandwidth efficiency. A typical web page is about 1/3 HTML and 2/3 images with respect to the bits required to represent the page. Hence, technologies that improve the bandwidth efficiency of image delivery are a logical first step. Static edge caching is one such technology. Static edge caches are deployed at locations near a concentration of users. The static edge cache sits between the users and the Internet. When a web object is requested for the first time, the cache forwards the request to the web server. The static edge cache sends the response to the user, and saves a copy. This copy is used to satisfy subsequent requests for the same object from other users, saving redundant roundtrips to the content provider. A significant limitation of static edge caching is that it is suitable primarily for static content that is of common interest to a large number of users. A good example of static content is a news image that is common to all users.
Static edge caching is not useful for dynamically generated or personalized content, leaving open an unfilled market need. A good example of dynamically generated content is a web page containing secure transactions, stock quotes, billing information or inventory controls.
THE SHIFTING BOTTLENECK
Prior to the widespread use of static edge caches, the outbound data traffic at a company's data center reflected the same composition as that of the content delivered. However, after the widespread deployment of static edge caches, images are served by the company only once per static edge cache, while dynamically generated HTML is served once per visit. This results in a shift in the composition of outbound traffic as shown below. In brief, the overwhelming majority of the outbound traffic today is dynamically generated HTML. This shift in the composition also shifts the bottleneck in web content delivery in that much of the download delay experienced by the user is due to the time required to generate and deliver the dynamically generated HTML.
EASING THE BOTTLENECK
Several technologies are emerging to address the acceleration of dynamically generated web content, in the form of both products and services.
Many of the products are focused on accelerating the generation of content, i.e., accelerating the performance of the web-server in creating the content. Typical of such products is the so-called dynamic cache that can augment server hardware by caching dynamically generated content. Since dynamic content has unpredictable lifetimes, the dynamic cache must closely track the freshness of cached content. As a consequence, dynamic caches must be tightly integrated with the application and the database, leading to higher integration costs. In light of this, adding server hardware is often more cost-effective in accelerating the generation of dynamic content.
Services for accelerating the delivery of dynamic content are typically offered as overlays to the Internet via Content Delivery Networks (CDNs). While CDN's are very effective in exploiting caching for the delivery of static content, they require complete reauthoring of the content to handle dynamic content. Furthermore, CDN's cannot effectively handle content that must be delivered SSL-encrypted end-to-end. Consequently, CDN's cannot transparently accelerate dynamic content and fail to address the 'last mile.'
In contrast to dynamic caches and CDNs, Axis Data transparently accelerates the delivery of dynamic content, secure transactions and embedded objects. Specifically, Axis Data's Concentrator product eliminates redundant traffic on the network by computing and transmitting only the changes that occur in a web page between successive downloads of the same or similar pages. The Concentrator is installed as a transparent proxy between the web server and the firewall. Using a technique known as Delta Optimization, the first time a user requests a document, the Concentrator stores a copy of the document in addition to sending the user a copy. At subsequent requests for the same page, the Concentrator receives the web server's response, compares the response to the stored copy from the prior visit, and computes a modified response comprising only the differences between the two documents. The Concentrator then sends the user the modified response, thereby eliminating unnecessary network traffic. Similarly for embedded objects, the Concentrator leverages another innovative technique known as FlashToo object acceleration. FlashToo allows the Concentrator to reduce both upstream and downstream traffic and eliminate network delays associated with web objects.
The Concentrator is transparent end-to-end, and requires no changes to the content, content server, or the end-user's browser.