Enterprise WAN Visibility: How to Ensure Performance of Carrier Ethernet Services

By Greg Gum  |  Posted 2010-01-13

Enterprise WAN Visibility: How to Ensure Performance of Carrier Ethernet Services

As enterprises expand operations globally, borders between countries continue to blur. Business is being conducted cooperatively from a constellation of offices, and IT managers' challenges are growing. These challenges are growing beyond LAN management into the need to expand service applications to a mostly dynamic, geographically-dispersed client base on a global scale. Driven by the need for this expansion, and supported by Ethernet's rapid evolution from a simple LAN connectivity tool into a carrier-grade technology (and the de facto protocol for next-generation broadband networks), enterprise IT infrastructures have recently undergone a major transformation.

Today, as an important component of IT infrastructure, Ethernet is becoming the foundation for WANs, with mission-critical links for site-to-site connectivity and Internet access. Carrier Ethernet, with its differentiated service capabilities, higher bandwidth and compelling economics (compared to traditional WAN technologies), is further motivating enterprises to use Ethernet as the tool of choice. Enterprises are using it to expand their IP-based service applications between geographically-dispersed locations, while migrating to higher speed WAN connectivity.

This extension enhances multisite collaboration and boosts productivity, but also increases the burden on IT managers since the infrastructure investment to secure and optimize the LAN must now traverse the WAN-efficiently and cost-effectively regardless of the user locations. As a result, real-time network performance monitoring is an important tool for troubleshooting network problems and assuring network service reliability. Carrier Ethernet enables service providers to have greater control over their networks which, in turn, enables IT managers to have greater network predictability.

Recent statistics and case studies from the Metro Ethernet Forum (MEF) have shown that enterprises can save significantly by moving toward Carrier Ethernet services. The primary motivation behind the departure from legacy frame relay, Asynchronous Transfer Mode (ATM) and Synchronous Optical Network (SONET)/Synchronous Digital Hierarchy (SDH) business services is economics, given that the typical 3-year costs of legacy solutions such as Private Line and frame relay are five to six times more expensive than Carrier Ethernet E-Line services.

Operations, Administration and Maintenance

Operations, Administration and Maintenance

While Carrier Ethernet is economically attractive, the focus of innovation has shifted to the area of Operations, Administration and Maintenance (OA&M), given the lack of new tools and software to manage demanding new service applications. Due to the dynamic nature of new real-time and near real-time bandwidth-intensive, IP-based multimedia applications, new OA&M management systems are expected to support and optimize the feature functionalities that are needed for better service performance and network efficiencies.

They're expected to do this while minimizing operating expenditure (OPEX) impacts to carriers and IT managers. Recent data from the FCC suggests that over 77.1 percent of an operator's service revenues are being spent on OPEX alone. This may explain the increasing trend toward outsourced management activities such as hosted services or software as a service (SAAS), given the large amount of technical and economical challenges required to maintain these services.

Thus, options available to large, in-house IT management departments are being scrutinized by corporations because, as service applications become more complex and OPEX heavy, pressure mounts to lower corporate operating costs-with managed services being viewed as one of the means to control expenses and head count within organizations.

Although outsourcing network management may still be the preferred option for some, there is clearly a need for new, advanced OA&M solutions to place the control back in the hands of the IT Managers-by decreasing IT's burden of network management and lowering the dominating factors of OPEX as the complexity and exponential growth of IP-based applications continue.

According to a recent FCC study, a typical enterprise's LAN administration costs constitute about five percent of the overall corporate OPEX model. However, with WAN services, this may grow significantly-especially since WAN service costs are monthly, recurring charges (which may also include premium services due to the service-level agreement (SLA) guarantees that are required to maintain high availability performance metrics). Therefore, IT managers must ensure the performance and efficiency aspects of the delivered services, and do so without further increasing their monthly operational budgets.

Over the past few decades, performance-monitoring tools for predicting network problems and enhancing network service reliability have been developed and implemented successfully for time-division multiplexing (TDM)-based networks. While the service space is changing, many of the problems and challenges of packet-based Ethernet networks are similar to those of the TDM era. Enterprises are demanding the same monitoring tools for IP networks that they used for TDM networks.



Standardization activities have made significant progress for accomplishing end-to-end visibility for Carrier Ethernet services. The IEEE 802.1ag and 802.3-2005 Clause 57 (formerly 802.3ah), ITU-T Y.1731, RFC2544 and MEF 17 specifications are just some of the standardized OA&M methodologies. Recent trends in the industry have led to the adoption of field-hardened, TDM-era methodologies leveraged into standard Ethernet OA&M-based tools that are integrated within network elements-empowering IT managers to monitor and manage networks on an end-to-end basis.

Since many Carrier Ethernet service networks are still best effort and may not provide detailed network metrics reports, the motivation behind adding monitoring capabilities became clearly visible through customer-reported data, indicating significant savings and rapid ROI when implementing advanced OA&M SLA monitoring:

1. Proactive network monitoring can result in up to 35 percent reduction in trouble tickets, 50 percent reduction in truck rolls and 70 percent reduction in on-site repair labor costs.

2. Integrated testing capabilities enable up to 50 percent reduction in external test setups and services.

3. In-field network equipment upgradeability to add end-to-end network visibility and OA&M into any network; eliminates network overhauling via forklift upgrades to add OA&M.

4. Integrated threshold-crossing alarms, switchover and other self-healing events enable 99.999% ("five nines") network and service reliability.

Carrier Ethernet has come a long way to become the de facto technology of broadband networks, and related monitoring systems have also improved over time. Real-time Ethernet WAN monitoring and optimization capabilities can provide up to 300 percent service performance improvement and up to 50 percent bandwidth efficiency uplifts.

While IT managers expand complex IP applications to their clients, regardless of the geographic location, technical challenges and cost-cutting pressures have prompted adding visibility into network health, service performance and verification of SLAs to ensure that precious OPEX and CAPEX are being appropriately managed. Since many Carrier Ethernet services are still best effort services, IT managers need reliable solutions to maintain network visibility, as well as higher layer services in real-time to ensure high levels of user satisfaction.

Further, using Ethernet monitoring tools saves precious OPEX and enables IT to proactively manage WAN connections (and their service providers' SLA guarantees) more efficiently and with less IT intervention through integrated testing capabilities.

The good news is that IT managers finally have new tools that provide the necessary WAN SLA network metrics to directly manage service provider-based Ethernet services. This ensures that user applications are reliably transported across the WAN.

Greg Gum is the Chief Marketing Officer at ANDA Networks. Greg brings over 20 years of experience in the telecommunications and networking industry, having served in various product management, marketing, corporate investment and business development roles at both startup and Fortune 500 companies. He can be reached at ggum@andanetworks.com.

Rocket Fuel