Andy Huckridge

Preamble

In April 2012, Gartner released an industry landscape which covered Network monitoring and defined a new term “Network Packet Brokers” – this paper will look at what they do, their advantages and why you should deploy one within your network.

Introduction

Carriers and Operators constantly face a challenge of containing costs while responding to the rapidly changing business environments, especially when it comes to maintaining high visibility and compliance in their networks.  The need to scale and remain compliant and secure, are at the forefront of their requirements. Unique for carriers and operators is the additional issue of ever increasing subscriber interest in, and revenue dependence on, new mobile services – raises concerns about spotty network performance. These require an increase in the quantity and variety of network analysis tools to attach to network links which can drive Capex and Opex out-of-control.

What is needed is a ‘layer of intelligence’ to provide scalable visibility and control of packet brokering – this is called network intelligence optimization, or intelligently optimizing (capturing, reviewing, evaluating and reporting) what is going on within the network. This important additional layer reduces both Capex and Opex and provides complete visibility across the network. By proactively copying, forwarding, and redirecting captured traffic in real-time, Network Operators can achieve higher service availability, lower labor costs for network analysis, lower analysis costs overall, and greater subscriber and revenue protection. In order to achieve the greatest level of return on investment (ROI) these tools are needed to optimize and secure the flow of network analysis to any operations center, on any network, anywhere – helping the Operator fuel a sustainable competitive advantage and guarantee future operating success.

Companies should consider an ROI evaluation that includes financial analysis and compares the estimated total cost of ownership (TCO) over three years between the current monitoring environment and a proposed Network Packet Broker layer.  With this information, the Operator can better understand the potential direct and indirect savings to their organization, including how productivity will be enhanced as subscriber experience, service performance, and network maintainability are all improved. A system wide interconnected approach to deploying a monitoring / analysis measurement system will maximize return on new tool investment, and provide a greater return on existing investment.

A clear alternative which either reduces or preserves current levels of Capex and Opex, is to deploy a Network Packet Broker layer, thereby extending the life of existing tools even as the network capacity is upgraded, visibility of the network is increased system-wide, future tool expenses can be better managed / predicted and a centralized monitoring system realized. These advantages bring clear benefits to the operator top and bottom lines.

Whether the IT group is centralizing their network intelligence tools or the tools will remain distributed throughout the network – LAN, WAN, or across the Cloud – a Network Packet Broker layer will provide complete visibility to the tools. Network “blind spots” can be eliminated because there is no more SPAN Port contention as the tools access traffic through the Network Packet Brokering system. Captured and groomed traffic maintains a centralized view of the network, regardless of physical location of the tools.

The immediate impact of the unprecedented network visibility is simplification of the network architecture, and a significant reduction in number of tools and consequently the management overhead. True end-to-end troubleshooting is now possible, resulting in much reduced response time to outage and repair (e.g. MTTR).

The intelligence of the system filters and grooms traffic to dramatically improve the efficiency of the tools. Many tools are application-specific, which means they are only interested in certain types of IP traffic coming from certain parts of the network. Selective hardware-based filtering, high data burst buffers and session-aware load balancing ensure that tools receive only the specific traffic they need to see (e.g. from specific VLANs), and that no packets are lost to due to over subscription.

With a Network Packet Brokering layer between the tools and the network infrastructure, instead of a 1:1 ratio of tools to network links, network operations can monitor several links or the entire network with a single tool. This dramatically reduces the Capex needed to completely cover the network.  This will result in lower management overhead, shorter time to troubleshoot network anomalies and repair mean further reduction in operating costs, faster ROI, and the ability to meet SLAs.

Summary

Network Packet Brokers, when used as a system enable the following distinct advantages:

• Up to 120% Increase in Application Availability
• Up to 80% Reduction in Capital expenditures
• Up to 50% Reduction in Operational expenditures
• Lower Mean Time to Repair (MTTR)
• Allows Operators to become truly tool vendor agnostic
• Lower churn and higher subscriber QoE
• Allow operators to focus on operational efficiency & service differentiation
• Allow monitoring tools to be cross-connected between different topologies as needed: Fixed, 3G, 4G/LTE and VoD
• Preserve existing investment, defer new investments
• Day 1 ROI, reduce tool TCO

Andy Huckridge, M.Sc (Telecoms)

Network Packet Brokers are also known by these other industry names: Fabric Switch, Traffic Visibility Switch, Visibility Layer, Traffic Visibility Networking, Intelligent Tap, Aggregator / Aggregation, Network Monitoring Switch, Distributed Traffic Capture, Network Intelligence Optimization