REMOTE MANAGEMENT

From the April 2005  issue of Communications News

The need for remote management

The time required to identify, isolate and resolve a problem at remote sites can be five times longer.

by Jeff Schmitz

No longer are organizations utilizing pure hub-and-spoke network architectures. The days of a network manager just worrying about bandwidth and connectivity to and from a headquarters location are numbered. Owing to the growing distributed nature of employees at remote sites–and the applications that now must run over the network to serve these individuals–the network manager’s world is in flux.

Additionally, there is growing pressure on the network. Traditional voice and data worlds are colliding with the widespread deployment of voice over IP (VoIP). Enterprises are increasingly tied to applications that traverse the network infrastructure. The rapid migration of both end-users and applications to remote locations is compounding the pressure on the network manager.

The responsibilities and requirements for network managers are growing more complex because of the confluence of three key trends: a growing remote workforce, the nature of distributed applications and the increased performance requirements being placed on the network.

A recent study indicates that 86% of workers do not work at a corporate headquarters location. Key drivers of this shift include multiple data centers, cost containment and flexible employment options. One of the greatest challenges for remote workers is the lack of onsite IT support staff for monitoring and troubleshooting network- and application-related performance issues.

As workforces continue to become more remote, the distribution of applications and servers is mushrooming. With deployment of solutions such as VoIP and new Web-based applications, the need for a single location to house applications and servers is no longer necessary. A single headquarters location can become a liability as a single point of failure, or adding unnecessary overhead and delay for mission-critical and time-sensitive applications.

As the workforce and applications become more distributed and the importance of applications running over the network grows, the performance and reliability of a remote location becomes as important as the headquarters location.

Application management needed
Studies have shown the amount of time needed to identify, isolate and resolve a problem at a remote location can be five times or more longer than a similar problem at a headquarters location. This disparity can be tied to lack of support staff, and the lack of tools and visibility at the remote sites.

In such a situation, a cost-effective remote network and application-management solution can improve both network and application performance. There are three pillars of such solutions: cost-effective remote visibility, network management and application management.

Many enterprises try to manage remote locations with a single head-end tool. Such a tool may identify there is an outage at a location, but does not provide the requisite deep remote visibility that is required to quickly isolate and resolve the problem. Deploying multiple head-end analysis tools to all remote locations typically is not an option due to limited IT budgets and staff.

The key is to find a system that can cost-effectively scale across remote locations while providing deep analysis and visibility for each location. A tool that looks network-wide but only polls every 10 or 15 minutes does not provide sufficient granularity for troubleshooting a problem. A head-end appliance provides deep visibility into a single location but does not allow for extensive remote views.

The network-management component has three distinct aspects that are increasing in complexity: the number of remote users, the growth in applications and the type of network/bandwidth. While the number of remote users is growing, the number of applications, along with their importance, is growing even more rapidly.

The growth in applications and remote users is also impacting the type of network and bandwidth required. Enterprises are migrating to IP virtual private networks with dedicated connectivity or DSL, or are moving to multiprotocol label switching-based networks with class-of-service (CoS) prioritization and any-to-any connectivity. A solution should be able to handle multiple network technologies to optimize performance, especially for remote locations.

An effective network-management tool should also be able to monitor and troubleshoot network components–ranging from physical access to ports to end-to-end virtual circuits to CoS settings. In addition, it should be able to troubleshoot problems as they occur, as well as being capable of scrolling back over time to identify intermittent issues or trends within the infrastructure. Finally, a granular view into network utilization is critical to optimizing.

The biggest recent change for network managers has been the need for application visibility. The challenge for many enterprises is if the organization breaks network and application infrastructure into two separate groups with different management, tools and budgets. The challenge is made more pronounced because any degradation of application performance is usually first submitted to the network manager, who must try to identify, isolate and resolve the issue.

Without tools and visibility for the application, the network manager typically tests the network components (e.g., CPE, transport) and works with the service provider to ensure service levels have been met. If the degradation is deemed not to be network-related, the problem is handed off to the application group–a way of ensuring the performance of the business-critical application.

A successful solution should understand the impacts of the application to the network and the network to the application. For example, poor performance of the application might be due to poor service-level parameters from the network. Conversely, poor network performance may be caused by unplanned applications using excessive bandwidth. Visibility into the end-user performance of the application and how it traverses the network is crucial to managing and troubleshooting network and application performance.

Enterprises should take a holistic view of the network and applications, and how they interact. For the remote sites, a lack of staff and lack of visibility equals a lack of control. A cost-effective remote network and application-management solution can help reduce the risk of poor performance and profitability as applications and users move from a central site to a more dispersed environment.

For more information from Visual Networks:
www.rsleads.com/504cn-257

Jeff Schmitz is senior vice president, marketing and product management, for Visual Networks, Rockville, Md.