Private Network Technologies Private Network Technologies in Software

Printer Code 128B in Software Private Network Technologies Private Network Technologies

TABLE 148 There are no standards for acceptable levels of the Frame Relay performance parameters listed below What is acceptable for one network may not be on another because of the data rate of the line, the amount and type of data traffic, and many other factors However, these guidelines can be applied for many typical Frame Relay implementations

Used to Indicate User data rates relative to the committed information rate (CIR) Data marked Discard Eligible Congestion in the Frame Relay network Device congestion Misconfigured routers, nodes or applications Consumption of interconnect device bandwidth by applications Consumption of network (transmission media) bandwidth by application Efficiency of networked applications

Guidelines <100% CIR utilization Bursts above the CIR are eligible for discard by the network Bursts above CIR with FECN or BECN congestion may indicate packet loss Device dependent Network-dependent (generally <20 to 30 per second) Network and application dependent Network and application dependent

Application dependent (generally, larger frames are more networkefficient than smaller ones) Network dependent Network dependent

Consumption of interconnect device bandwidth by node Consumption of network (transmission media) bandwidth by node

FECN (Forward Explicit Congestion Notification) is a frame relay flow control flag bit used to notify the receiving node that there is incoming network congestion BECN (Backward Explicit Congestion Notification) is a frame relay flow control flag bit used to notify the sending node (source node) that there is network congestion on the outbound path The suggested response is to reduce the frame rate into the network

Downloaded from Digital Engineering Library @ McGraw-Hill (wwwdigitalengineeringlibrarycom) Copyright 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website

Downloaded from Digital Engineering Library @ McGraw-Hill (wwwdigitalengineeringlibrarycom) Copyright 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website

Introduction This chapter examines performance testing in private networks, including customer premise local area networks (LANs) and wide area internetwork (WAN) connections It includes criteria for evaluating network performance, factors that affect network performance, network performance metrics, tools for monitoring network performance, and baselining and benchmarking techniques for characterizing network performance

Network Performance Criteria Network performance can mean different things to different people A user on a networked computing system ideally should be completely unaware that the network exists Network performance from a user s perspective can be thought of as the degree to which the network is completely invisible, as if the user were directly connected to any resource he or she chooses to access Network performance can be defined as reliability, availability, data throughput, error rate, response time, application performance, or in many other different ways As each network is unique, so too are the criteria that define performance for each individual network It is important for each network manager to understand what constitutes good performance for the network being managed, so that results can be measured and compared against a goal

One very basic measure of network performance is reliability: Can the network be depended upon Reliability is a perception held by the network s users It is based

Downloaded from Digital Engineering Library @ McGraw-Hill (wwwdigitalengineeringlibrarycom) Copyright 2004 The McGraw-Hill Companies All rights reserved Any use is subject to the Terms of Use as given at the website

upon history: past downtime, application performance, and response time While each of these factors is quantifiable, the interpretation of what is or is not a reliable network depends on the situation Every network is different The real value of downtime, application performance, and response time measurements comes with regular network baselining, where a history of these measurements can be examined for changes and for trend analysis The baselining process is examined in more detail later in this chapter Networks that are perceived by users as unreliable will not be used, at least not to the extent that they otherwise would be User productivity will suffer as a result