Chapter 10 Monitoring and Troubleshooting 10.1 Network Monitoring Network MonitoringNetwork Monitoring 10.2 Troubleshooting Troubleshooting.

Chapter 10 Monitoring and Chapter 10 Monitoring and TroubleshootingTroubleshooting

10.1 10.1 Network Monitoring

10.2 10.2 Troubleshooting

Chapter 10 Monitor & Troubleshoot

10.1 Network Monitoring

Network monitoring ensures a network to function continuously Aspects include

Server performanceNetwork statisticsHard disk performanceMemory usage CPU utilisation


10.1.1 Server Performance

Server performance includes number of queued commands

queued commands = commands awaiting executiontoo high → a bottleneck in the server

• use additional server

• use multiple processors

quality of server sessionsserver session records server-client connection How is the connection broken?

• normal logoff, error, or server timeoutserver timeout → server overloaded


10.1.2 Network Statistics

Network statistics include error rate

number of collisions per second• re-transmission after collision

• waiting time will increase greatly

network utilisation

Saturated• utilisation ~100%

• →more collisions

Actual bits transmitted per second

Bandwidth


10.1.3 Hard Disk Performance

Hard disk performance includes disk space remaining throughput of the hard disk

Throughput = amount of data transferred per secondthe speed at which requests are serviced.

number of queued requestsbusy or notToo high → hard disk overloaded


10.1.4 Memory Usage

Memory usage ← rate of memory swapping Swapping

data not in use are moved from the main memory onto the hard disk to make room for new data

Too many swapping → system slows down


10.1.5 CPU Utilisation

CPU utilisation average percentage of CPU activity high (> 90%) → overloaded or, caused by

low memoryimproper configuration


10.2 Troubleshooting

Common network problems: response too slow unable to detect the network from a workstation unable to login unable to use resources of the network unable to connect to the Internet unable to execute certain programs



Troubleshooting to diagnose and fix problems

The structured steps are establishing the symptoms identifying the affected area establishing what has changed selecting the probable cause implementing the solution testing the results documenting the solution



10.2.1 Establishing the symptoms

Symptoms are information about a problem may come from

users• 1. when the problem occurs

• 2. what the user is doing when the problem occurred

• 3. how often the problem occurs

• 4. what has been done to try to correct the problem

• 5. what the system has changed recentlyerror messages or event logs



10.2.2 Identifying the affected area

scope a single workstation a network segment (controlled by a hub/bridge/switch) the entire network

involves network infrastructure• malfunctioning hubs or switches

• insufficient bandwidth for the networkhighest priority



10.2.3 Establishing what has changed

A network suddenly fails → some changes Consider:

new software or hardware added? positions of hardware on the network changed?

e.g. Moved workstation → cabling related issue



10.2.4 Selecting the probable cause

Selecting the probable cause requires isolating the problem tools

If a user is unable to log onto the network, probable causes:

invalid logonfaulty cablebad network cardthe account has been locked




A. Isolating the problem Techniques

removing cables one by oneshutting down services one after the otherpulling out affected system

• check whether it can work as a standalone machinerebooting the network

• last resort




B. Tools Cable tester

a device used in checking a cable

Event viewer a utility program that stores

• security threats

– e.g. failed log on or failed data access

• system activities

• application activities

– tell how well the system is running







B. Tools (cont’) Protocol Analyzer / Packet Sniffer

is hardware or softwareverifies the working of a protocolcaptures each packet passing over a network

• and decodes its content

Note: packet sniffers may be used by hackers to intercept data






10.2.5 Implementing the solution

Common solutions: replacing with a new hardware or reconfiguring the server.

Note: 1. If powering off is required, plan

the time to power offmaking backup

2. Implement solution with one fix at a timeotherwise, difficult to tell what has corrected the problem and

why



10.2.6 Testing the results

Testing after implementing the solution to verify that the solution really addressed the problem



10.2.7 Documenting the solution

Documentation provides a quick reference for future include

what the problem washow (and when) the solution was implementedother solutions that have been tried out who implemented the solution

Chapter 10 Monitoring and Troubleshooting 10.1 Network Monitoring Network MonitoringNetwork Monitoring 10.2 Troubleshooting Troubleshooting.

Documents

network yunable

solution slide

event logs slide

hard disk xto

hard disk xthroughput

additional server

login yunable

internet yunable