Figure 1. The advantages of aggregation and filtering in an enterprise environment


 

Anyone who runs a business today knows that the economy, competition, regulations, and other forces have combined to create an unforgiving business environment. Enterprise executives in this environment struggle to achieve profitability, tasking every department to cut costs and increase productivity.

Given that nearly every commercial enterprise relies on its network for operations as well as revenue generation, it’s surprising that more enterprise executives have not heard the message from their IT departments-investments in network management technology are key to growing the overall bottom line.
 

Monitoring Tools

When times are good, it’s easy to overlook the more focused or “niche” elements that comprise a typical enterprise network infrastructure. Hardware such as servers, switches, and routers certainly get top-line visibility, but the devices that monitor their performance often don’t. However, today’s challenges provide extra motivation to look at monitoring devices.

A report last year by the industry analyst firm Frost & Sullivan adds evidence to this notion. The report states that while the total network performance monitoring and solutions market remained flat in 2009, it is expected to grow by 7.2 percent in 2010 to more than $1.3 billion. Such data would seem to support what we’ve heard from IT executives, namely that these devices are now becoming part of a standardized solution for enterprise data centers.

Among the tools being adopted are wide area network (WAN) analyzers. IT departments are purchasing them in order to gain insights about critical WAN links, network performance, and adherence to service-level agreements. Other devices being evaluated include application-aware analyzers, which eavesdrop on live communications between servers and clients. They can decode network and application-layer protocols, parse out client/server information, and determine error rates, thus providing an accurate picture of what is taking place on the fiber or wire.

The point that IT executives and data center managers need to continue to emphasize is that these tools are essential for maintaining a healthy network and meeting security and compliance regulations. This is underscored by the observation in the same Frost & Sullivan report, which states, that “Internet usage is lead[ing] to increased security fear among end users.” It is not unrealistic to conclude that having these tools more readily accessible and more easily applied to suspect network trouble spots will go a long way to alleviating such anxiety.
 

Packet Aggregation Switching

Given these concerns, it is understandable that packet aggregator switches have begun to attract more industry attention. These network devices combine aggregation and filtering technology with matrix-switching capabilities, enabling the electronic sharing of analysis tools-in turn, making it possible to reduce the overall investment in monitoring equipment. Some Apcon customers have reduced their investments by as much as 50 percent per data center, resulting in savings of more than $1.2 million in analysis tool expenditures and nearly $200,000 in annual maintenance fees.
 

Figure 2. An example of an interface that enables efficient, centralized management of network analysis equipment and allows IT managers to see where their monitoring devices are deployed

 

Such savings opportunities are important given that data center managers are now faced with the requirement to effectively monitor both 1G and 10G network segments-a trend that is spurring the need for additional investment in 10G analysis devices. The unfortunate reality, however, is that organizations have continued to rely largely on 1G devices due to budget constraints.

Packet aggregator switches offer a way to use the existing array of 1G tools to monitor both the 1G links as well as the 10G network infrastructure. The technology is growing in appeal because of advances such as higher port-density chassis, which eliminate the need to trunk and stack multiple devices together-a practice that introduces both the potential of multiple points of failure and bandwidth bottlenecks that can limit any-to-any connectivity.
 

Software

Enterprise executives are also taking note of recent developments in the software arena, particularly programs that enable centralized management of network analysis equipment. These programs are appealing for several bottom-line reasons. First, they speed and simplify the monitoring and managing of even the most complex networks. Also, they enable IT management to see which network monitoring devices are being used at maximum capacity. This, in turn, leads to better utilization of equipment and the proper allocation of CAPEX dollar

In one case, a major telecommunications service providers with five data centers located through the U.S. has deployed switching solutions and associated software to monitor the enterprise’s networks and facilitate connections to an array of analyzers and related tools for the purpose of troubleshooting and problem resolution.
 

Figure 3. A many-to-many monitoring approach for monitoring of 10G links with 1G tools

 

Recently, one of the data centers-a lights-out facility-experienced an alert that one of its significant customer-facing transaction applications was experiencing a performance slow-down. Upon receipt of the alert, an engineer in another data center was able to remotely access the lights-out facility and electronically isolate and solve the problem and bring the application back to full operation in approximately 90 minutes-all via an embedded software interface integrated with the switching solution.

By having this remote-access and software-based problem-solving capability, the enterprise avoided the typical process that would have required the issuance of multiple trouble tickets, the dispatch of a technician to the data center, and the task of physically connecting/de-connecting an analyzer tool to isolate the problem and resolve it. The 90-minute resolution process contrasts with the typical manual process-approximately 24 hours. It is estimated that savings to the enterprise amounted to hundreds of thousands of dollars if not more.
 

Conclusion

IT and data center management have a strong business case to make to C-level executives of any global enterprise. The organization needs to acknowledge that focused infrastructure elements such as monitoring tools are critical assets. In turn, IT managers who can prove-quantifiably-the impact of these devices will enhance their stature in the eyes of senior leadership. More important, their action will build recognition of and support for technology investments that promise significant efficiencies and cost-savings to the enterprise-as a result, enhancing the organization’s prospects for competitive advantage, profitability and growth.