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




Seven Web Load Balancers Score With Round-The-Clock Access

With ClusterCATS' SmartCluster set up, we were able to use the replication tools to help us distribute our Web server test files. Three methods are available. First, ClusterCATS can autodiscover new Web content and automatically update it. Second, Web administrators can use the ClusterCATS Explorer to drag and drop Web content from server to server or from any other Explorer window.

Finally, each server can publish or subscribe to another to constantly update information in a particular directory. ClusterCATS is the only product we tested that has this capability. With features like these, ClusterCATS may be better classified as a resource-management solution than a load balancer.

Each server in the SmartCluster can serve content without interaction from a network device or additional servers, so there is no load-balancing mechanism that requires a backup. SmartCluster servers are backed up by one or more alias servers--a Windows NT server with two NICs installed and the ClusterCATS server bound to the primary NIC.

The secondary NIC is then bound to back up addresses as needed when ClusterCATS servers go down. If a server failure is detected, the secondary NIC assumes the IP address until the downed server comes back. ClusterCATS detects the recovered server via the NT Event Log. We consider this completely unacceptable for an enterprise network. In the event the Alias server fails, the other ClusterCATS servers are incapable of assuming its traffic without another Alias server.

ClusterCATS handled our test traffic well, but its reliance on Windows NT and IIS 2.0+ makes it unacceptable for many sites. A Solaris version was in the early beta stages during our testing and it may be available as soon as August 1. ClusterCATS may be better suited for smaller workgroup-like installations; for enterprise network load balancing, look elsewhere.

Cisco Systems LocalDirector 410 1.6
Although there's nothing seriously wrong with LocalDirector, it falls short of the high standards set by its hardware brethren the WSD Pro and ServerIron. While configuration is simple and its commands are clear and easily understood, the absence of any GUI management and real-time monitoring utilities is indicative of a subpar load balancer. Performance was also lacking in our T3 and Fast Ethernet tests.

LocalDirector is most noted for its simplicity. It didn't take long for us to familiarize ourselves with the LocalDirector console interface. However, LocalDirector comes up short on features when stacked up against almost every other load balancer. It has no management utilities aside from telnet and serial access. Configuration information must be entered one line at a time, which quickly becomes tedious for larger Web sites, although we were able to streamline the process a bit with a cut-and-paste procedure using both telnet and serial connections.

The LocalDirector also acts as an Ethernet bridge that prohibits administrators from completely hiding their Web servers' real IP addresses. Failover is accomplished via a serial cable, so there's a limit on how far apart you can place redundant units.

LocalDirector has a couple of worthy features. Web servers that are part of Web clusters can be brought into service over a long period of time, thus avoiding a bombardment of new connections the second they are brought up.

This slow start gives administrators time to ensure that a new server is working correctly rather than having to instantly rely on it for general use. In addition, servers can be placed in and out of service quickly and easily. When a server is flagged as out of service, users who are already connected will not be kicked off, but new server connections will be blocked.

Gregory Yerxa can be reached at gyerxa.nwc.com.


Other Reviews
Holy Intruders!: IP-Based Security Auditing Tools
By Greg Shiply


Print This 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

Data Centers Gone Wild
February 22, 2010

NWC


Salary

Video