Upcoming Events

Executive conference

Cloud Connect March 16-18

Comprehensive thought leadership for executives, IT professionals and developers. Topics include: the ROI, cost and economics of on-demand computing; Migration strategies to move from on-premise to cloud-based IT; Vertical cloud specialization, tailoring features and architectures to specific applications, industries, and customer ecosystems

More Events »

Subscribe to Newsletter

  • Keep up with all of the latest news and analysis on the fast-moving IT industry with Network Computing newsletters.
Sign Up

  F E A T U R E

Web Content Switching

August 7, 2000
By Gregory Yerxa and James Hutchinson

A typical corporate Web site has become a complex grouping of specialized technology. Your site could easily include a firewall, a router, Layer 2/Layer 3 switches, load-balancing devices, cache servers and Web servers. You can choose to design it yourself or rely on any of the external hosting, content and management services popping up daily. How all these pieces are put together directly impacts your site's performance. If you want a high conversion rate on your Web site, you'd better be quick: End users lose patience when Web sites take longer than eight seconds to load.

One way to improve performance is to control how traffic enters your environment. Web content switches--also known as URL switches or Layer 7 switches--can provide you with the highest level of control over your incoming Web traffic. Web content switches look all the way into the HTTP header to make load-balancing decisions, rather than stopping at the TCP port number. By examining the HTTP header, these switches can make decisions on how individual Web pages and images get served from your site. This level of traffic control can be helpful if your Web servers are optimized for specific functions, such as image serving, SSL (Secure Sockets Layer) sessions or database transactions.

In our tests of Web content switches at the University of Wisconsin-Madison Real-World Labs®, we discovered that efficiently using this technology can improve site performance, take better advantage of available resources and better direct those resources to target audiences.

Two of the four products we reviewed--F5 Networks Big-IP Controller 3.1 and Intel Corp. NetStructure 7180 e-Commerce Director--are capable of accelerating SSL traffic in conjunction with traditional load-balancing and Web content switching. In recent months, SSL acceleration features have garnered much attention. SSL acceleration offers load-balancers the ability to act on traffic and information that is otherwise encrypted. With the Web session in the clear, load-balancers are free to access HTTP header information in the same manner as they do in unencrypted sessions. Web server performance is likely to improve with the SSL negotiation overhead off-loaded to a specialized device.

The winner of our Editor's Choice award is F5's Big-IP. Big-IP's straightforward C-style syntax and if-then-else rule logic made setting up and maintaining our test configuration a breeze. While the Alteon WebSystems Alteon 184, Foundry Networks ServerIronXL with Internet IronWare 7.0, and Intel NetStructure 7180 all functioned adequately, they lack the ease of use and streamlined configuration of Web switching rules that the Big-IP device possesses.

Is It Right for You?

In some cases, URL load-balancing devices can save you some money and maximize your equipment's capacity. If you support a site that both customers and employees use, for example, you may want to channel your employees to a slower server farm to keep resources free for your customers. Web hosting facilities also can use this technology to save on hardware costs, since they may move incoming customer traffic for many hosted sites to a common Web server farm easily. As customers' requirements grow, their traffic can be moved to dedicated Web server equipment, and the facility could, in turn, charge higher fees. But the network infrastructure stays the same, because the same Web content switch is controlling traffic.

These devices also can give you the ability to use all your available computing power. That's especially important if those start-up dollars are running thin. By switching traffic based on a URL request, you may be able to run some of the less popular areas of your site with less powerful Web gear and free up your high-end Web servers to handle most of the critical network Web application traffic. Management, syntax and rule logic play a critical role in using Web content switches for your everyday Web server needs. With adequate Web content-switching features, Web and network administrators can use this functionality to respond to flash crowds and intense changes in the resource needs of Web sites.

Added Complexity

With the current state of the technology, however, it's likely that you don't need Web content switching at all. It adds another level of complexity to Web sites that already are difficult to build, maintain and troubleshoot--especially if you derive direct business revenue from your site. Any technology that might hurt service without improving the customer experience is not required.

If your site is operating under heavy load, the performance hit you take by using Web content switches could cause more problems than it's worth. Just imagine the business impact if these advanced switching features are the sole cause of delaying your customers from getting to your site when they are trying to make purchases. Chances are very good those people will not be repeat visitors or customers.

Another issue to consider (as you would with any critical network component) is reliability. If one switch makes all traffic decisions for your Web site, some failover and redundancy features must be built in. Just as in the case of a core router, if that device fails and your site is designed to work exclusively by the URL rules you've established, you need to be able to recover as quickly as possible. Fortunately, redundant configurations abound. Some Web content switches exchange session information in real time to prevent existing sessions from losing persistence, and almost every vendor has some take on how to work with or without VRRP (Virtual Router Redundancy Protocol) by placing the switches in different locations on the network.

What If I Still Can't Make Up My Mind?

If you think you'll need Web content switching eventually--but not yet--not to worry. Many load-balancing equipment vendors support these features in their standard code releases. Your existing load-balancing equipment possibly has that feature set, but it was never turned on.

Finally, if you do have a strong business reason to implement Web content switching, make sure that your software and hardware Web architecture is ready to take advantage of it. Servers must be sized properly to handle the tasks they've been assigned, and performance and redundancy must remain the top priorities. Modifying how information flows through your Web site is not a trivial task. There needs to be synergy among all the technologies you use to run your Web environment. As with everything else in the Web world, nothing beats good old-fashioned planning.

The bottom-line cost of implementing Web content switching may hurt both your financial and your engineering resources. With the proper design when implementing Web content switching, however, you can reap the rewards that these features provide, gaining increased flexibility over content requests and traffic.


PAGE: 1 I 2 I 3 I 4 I 5 I 6 I 7 I NEXT PAGE
 

Best of the Web

Data deduplication: Declawing the clones

Data deduplication is emerging as a critically important new arrow in the storage administrator's quiver to answer hard questions about the increasing problem in storage growth costs.

Quick Read

Compression, Encryption, Deduplication, and Replication: Strange Bedfellows

One of the great ironies of storage technology is the inverse relationship between efficiency and security: Adding performance or reducing storage requirements almost always results in reducing the confidentiality, integrity, or availability of a system.

Quick Read

WAN Optimization Whitelists and Blacklists

Optimization is a fantastic way of saving money and creating really happy customers at the same time, but it doesn't work flawlessly for all applications.

Quick Read

WAN Optimization as a Managed Service: It's Not About the Cost

This insight examines how organizations outsourcing their WAN optimization initiatives to a third-party go about achieving their goals for application performance, reducing operational costs, and streamlining enterprise infrastructure.

Quick Read

  Sponsored Links

Premium Content

Next Generation Data Center, Delivered, November 17th
NWC


Salary

Video