3.1 Monitoring computer network

Monitoring a computer network involves the continuous observation of network components to ensure optimal performance and security. In Kenya, where institutions like county government offices and banks rely heavily on network uptime, effective monitoring prevents disruptions that could compromise service delivery. Network monitoring helps identify bottlenecks, unauthorized access, and hardware failures, enabling proactive maintenance rather than costly reactive fixes.

3.1.1 Introduction to computer network monitoring and maintenance

Network monitoring is a structured process that involves tracking the health, performance, and security of network infrastructure. Maintenance refers to the corrective and preventive actions taken based on monitoring insights to keep the network operational and secure.

Network Monitoring Objectives

• Performance Optimization: Monitoring bandwidth usage and latency helps ensure that network speed meets organizational needs, such as smooth transaction processing in SACCOs.
• Fault Detection: Early identification of link failures or device malfunctions reduces downtime at institutions like hospitals where network availability is critical.
• Security Surveillance: Monitoring unusual traffic patterns helps detect cyber threats, which is vital for safeguarding data at insurance firms.
• Capacity Planning: Tracking network usage trends enables ICT technicians to recommend upgrades before congestion occurs, supporting growing universities efficiently.
• Compliance Assurance: Ensuring that network operations comply with regulatory standards like the Data Protection Act protects county government offices from legal penalties.

Network Maintenance Types

• Preventive Maintenance: Scheduled tasks such as firmware updates and hardware inspections prevent failures and extend device lifespan.
• Corrective Maintenance: Actions taken to repair or replace faulty components identified through monitoring.
• Adaptive Maintenance: Adjustments made to network configurations in response to changing organizational requirements.
• Predictive Maintenance: Using analytics from monitoring data to forecast and prevent potential failures.
• Emergency Maintenance: Immediate interventions during critical faults to restore network functionality rapidly.

Importance of Network Monitoring and Maintenance

• Enhances network reliability and reduces service interruptions.
• Protects organizational data and maintains confidentiality.
• Supports efficient resource allocation and cost management.
• Improves user experience by minimizing latency and downtime.
• Facilitates compliance with industry and government regulations.

Monitoring techniques

Technicians combine several techniques to cover the whole network, since no single method shows every fault.

• Manual Inspection: Periodic physical checks by technicians to verify device condition and cable connections.
• Remote Monitoring Tools: Software that reports device health metrics, often used by SACCO IT teams to oversee multiple branches.
• Log Analysis: Reviewing device logs for hardware errors or warnings.
• Alert Systems: Configured to notify technicians via SMS or email when physical device anomalies occur.
• Scheduled Reports: Regular summaries of utilisation, errors and uptime allow trends to be compared month by month and support maintenance planning.

3.1.2 Computer Network Monitoring Physical Tools

Effective network monitoring requires physical tools that enable ICT technicians to assess and troubleshoot hardware and connectivity issues. These tools are essential for on-site inspections and diagnostics in environments such as retail businesses and county offices. The table below summarises the tools used in this unit; the cable tester, crimping tool and stripping tool are explained in the numbered sections that follow, and the other instruments are described straight after the table.

Other physical monitoring tools

Besides the cable tester, crimping tool and stripping tool, technicians use other instruments to check electrical, optical and termination quality.

• Network multimeter: A multimeter measures voltage, current and resistance and helps diagnose power issues in switches or routers. It is useful in network rooms where the power supply may fluctuate. A technician can also use it to check continuity of a cable and to measure the output of a power adapter. Always set the correct range and follow electrical safety rules before measuring mains voltage.
• Optical power meter: An optical power meter measures the strength of the light signal in a fibre optic cable, normally in decibels. It is essential for verifying the quality of fibre optic links used in county government WANs, ensuring minimal signal loss. The reading is compared with the expected power level of the link, and a low value indicates a dirty connector, a bend or a damaged fibre.
• Punch down tool: A punch down tool pushes the conductors of a cable into the contacts of a patch panel, wall outlet or termination block, and trims the excess wire. It is used to secure cables on patch panels, improving connection reliability in office networks. A good punch down seats each wire fully and keeps the untwisted length as short as possible.

3.1.2.1 Cable testers

A cable tester is a handheld instrument that sends a signal through each conductor of a network cable and shows whether the wires are connected correctly. It is used to detect breaks, shorts, and miswiring in network cables before a cable is placed into service or when a user reports a dead network point. Technicians at universities often use this tool to ensure proper cabling in computer labs.

A basic tester has a main unit and a remote unit. Each end of the cable is plugged into one unit, and LEDs for pins 1 to 8 light in sequence if the wiring is correct. Advanced testers, often called certifiers, also measure cable length, signal loss and interference, and report a pass or fail against a chosen cabling standard.

Faults a cable tester can reveal

• Open circuit: One or more wires are broken or not seated in the connector, so the matching LED does not light. This often happens when a wire was not pushed fully to the end of the RJ45 plug.
• Short circuit: Two or more wires touch each other, so several LEDs light together or at the wrong time.
• Reversed pairs: Wires are connected in the wrong positions at one end, for example pin 1 joined to pin 2. The LED order on the remote unit does not match the main unit.
• Crossed wiring: A straight-through cable has been wired as a crossover, or the two ends use different colour standards. The tester shows swapped LED positions.
• Split pair: The pin numbers match but the wires belong to different twisted pairs, which causes high interference. Only better testers detect this fault, so a basic tester may still show a pass.
• Excess length: A copper Ethernet channel should not exceed 100 m in total. A tester that measures length shows when a run is too long.

Procedure for testing a patch cable

1. Inspect the cable and both connectors for crushed jackets, bent pins or cracked plugs, because visible damage already explains many faults.
2. Disconnect the cable from the switch, patch panel or computer so that no live equipment is connected during the test.
3. Plug one end into the main unit of the tester and the other end into the remote unit.
4. Switch the tester on and select the automatic or wire-map test mode.
5. Watch the LEDs for pins 1 to 8, or read the result on the screen, and check that they light in order with no gaps or swaps.
6. Record the result against the cable label, noting any open, short or crossed pins.
7. If the cable fails, cut off the faulty connector, re-terminate it and test again.
8. Label the cable as tested and return it to service only after a clean pass.

3.1.2.2 Crimping tool

A crimping tool fixes an RJ45 connector onto the end of a twisted-pair cable by pressing the plug's metal contacts into the copper conductors. It allows technicians to create custom-length cables or repair damaged connectors, a common task in retail outlets upgrading network layouts. A poor crimp is one of the most frequent causes of intermittent connection faults, so the tool and the technique both matter.

A typical crimper has a die shaped for RJ45 plugs, a ratchet that forces a full squeeze before release, and often a built-in cutter. Some models also have a stripping blade. The crimp is always followed by a cable test.

Procedure for terminating a cable with an RJ45 connector

1. Cut the cable end square with the cutter so that the conductors are level and undamaged.
2. Strip about 25 mm of the outer jacket using the stripping tool, taking care not to nick the wire insulation.
3. Untwist the four pairs and straighten the wires between your fingers.
4. Arrange the wires in the chosen standard. For T568B the order from pin 1 to pin 8 is white/orange, orange, white/green, blue, white/blue, green, white/brown, brown.
5. Trim the wires level so that about 12 mm remain outside the jacket, then slide them into the connector with the clip facing down until every wire touches the end of the plug.
6. Check that the outer jacket reaches inside the plug, so that the strain relief can grip it.
7. Insert the plug into the correct die of the crimper and squeeze fully until the ratchet releases.
8. Use the same colour standard at the other end for a straight-through cable, then test the cable with a cable tester.

Good crimping practice

• Use the same standard at both ends: Mixing standards creates a crossover cable that will not work between a computer and a switch.
• Keep untwisted length short: Untwisting more wire than needed increases interference and reduces speed.
• Match the plug to the cable: Plugs for solid cable and stranded cable differ, and the wrong type gives poor contact.
• Do not reuse a crimped plug: Once the contacts are pressed in, the plug cannot be reliably crimped again.
• Test every cable: A cable that looks correct can still have an open or crossed wire.

3.1.2.3 Stripping tool

A stripping tool removes the outer jacket of a network cable so that the inner twisted pairs can be reached for crimping or punching down. Its blade depth is set to cut only the jacket, not the insulation of the wires beneath. A nicked wire may work at first and then fail later, which makes careful stripping an important part of reliable cabling in offices and computer labs.

Common types include a rotary stripper that is turned around the cable, and a combined crimper and stripper with a fixed blade slot. A cable with a thick jacket, such as outdoor cable, may need a knife or a cable jacket remover.

Procedure for stripping a UTP cable

1. Select a stripper suited to round UTP cable and check that the blade is clean and sharp.
2. Hold the cable end and place it in the stripper so that about 25 mm of cable extends past the blade.
3. Squeeze lightly and rotate the tool once or twice around the cable, or draw the cable through the slot, to score the jacket.
4. Pull off the cut section of jacket, and do not twist it sharply, as this can damage the wires.
5. Inspect the exposed pairs for nicks or cuts. If any insulation is cut, cut the cable back and strip again.
6. Remove any pull string or separator and separate the pairs ready for terminating.

Safety and care of the tool

• Keep blades sharp: A dull blade needs more force and tends to slip, which can cut fingers or damage the cable.
• Cut away from the body: Always strip with the blade directed away from your hands.
• Adjust for the cable: A blade set too deep cuts the wire insulation, and one set too shallow leaves the jacket attached.
• Store the tool dry and clean: Moisture and dirt cause rust, which damages the blade edge.
• Never strip live cables: Cables connected to powered equipment, such as Power over Ethernet devices, should be disconnected first.

3.1.3 Physical Networking Device Status Monitoring

Monitoring the physical status of networking devices such as switches, routers, and access points is vital to maintain network integrity. Physical indicators provide immediate insights into device health and connectivity status, enabling prompt response to issues.

Environmental monitoring of temperature and cooling

Heat and humidity shorten the life of network equipment, so sensors that track humidity and temperature around devices help prevent environmental damage. Switches and routers in a small cabinet or an unventilated room are especially at risk.

• Temperature sensors: Overheating signs trigger alerts. In data centres at universities, temperature monitoring prevents hardware damage.
• Fan operation: Proper cooling is essential. Failed fans in a network room can cause equipment failure, so listen for noise changes and check that fans turn and that vents are not blocked by dust or objects.
• Ventilation: Cabinets should have clear airflow, and equipment should not be stacked tightly with no gaps.
• Humidity: Very damp air causes corrosion, while very dry air raises the risk of static discharge.
• Cleaning: Dust filters and vents should be cleaned on a schedule, with the equipment powered down where the manual requires it.

3.1.3.1 Port and interface

A port is the physical socket on a switch, router or access point where a cable connects, and an interface is the logical connection the device uses for that port. Monitoring them shows whether a link is up, at what speed it is working and whether it is producing errors. Most devices have LEDs that indicate power, link activity, and error states. For example, a blinking green light on a switch port in a bank indicates active data transmission.

What to monitor on ports and interfaces

• Link status: A steady link light shows that the port has detected a device at the other end. No light means the port is down, the cable is faulty or the connected device is off.
• Activity: A blinking light indicates traffic. A port that never blinks may be unused, or the connected computer may have a fault.
• Speed and duplex: Ports negotiate a speed such as 100 Mbps or 1 Gbps. A port running slower than expected often points to a damaged cable or a mismatched setting.
• Error counters: Management software and the device's own interface can show counts of errors, dropped packets and collisions. Rising counts point to a failing cable, connector or network card.
• Utilisation: The percentage of port capacity in use shows which ports are heavily loaded, such as an uplink between switches.
• Unused ports: Ports that are not in use should be identified and recorded, since an unknown active port can be a security risk.

Procedure for checking a switch port

1. Identify the port from the patch panel label and the switch port number.
2. Look at the port LEDs and note whether they show link, activity or a fault colour. The meaning of colours differs between makes, so refer to the device manual.
3. Connect to the switch management interface and view the status of the same port.
4. Read the speed, duplex and error counters and compare them with previous readings.
5. Swap the patch cord with a known good cord if the port shows errors or no link.
6. Test the original cord with a cable tester if the problem moves with the cord.
7. Record the findings and any action taken in the maintenance log.

3.1.3.2 Cable and connection

Cables and their connections are the most exposed part of a network and the source of many faults. Regular visual checks identify loose cables, corrosion, or physical wear, common in outdoor network points in agricultural cooperatives. Monitoring them means checking both the cable itself and the places where it joins other equipment: connectors, wall outlets, patch panels and device ports.

What to check on cables and connections

• Connector seating: Plugs must click firmly into ports. A loose plug causes intermittent loss of connection.
• Cable condition: Look for crushed, cut, kinked or chewed jackets, and for cables pulled tight across doorways or under chairs.
• Bend radius: Sharp bends damage the twisted pairs. As a guide, UTP cable should not be bent tighter than about four times its diameter.
• Labelling: Both ends of each cable and every patch panel port should be labelled, so that a technician can trace faults quickly.
• Corrosion and moisture: Metal contacts that are green, dull or wet cause poor connection, especially in outdoor or humid locations.
• Cable length: A copper Ethernet channel should not exceed 100 m. Longer runs cause weak signals and slow or unstable links.
• Cable management: Cables should be tidily dressed with ties or trays, and kept away from power cables to reduce interference.

Procedure for inspecting a cabling run

1. Start at the wall outlet and check the faceplate, the socket and the patch cord for damage.
2. Follow the visible cable route and look for crushing, tight bends or exposed sections.
3. Check the patch panel for loose or unlabelled cables and for punched-down conductors that have lifted.
4. Reseat each connector by removing it and pushing it back until it clicks.
5. Test the cable with a cable tester and compare the result with the previous record.
6. Replace or re-terminate any damaged cable or connector, and test again.
7. Update the cable records and labels.

3.1.3.3 Power supply

A network device cannot work without stable power, so the power supply is a key item to monitor. Power supply units (PSUs) convert mains electricity to the low voltages the device needs, and monitoring PSU status ensures uninterrupted power. County offices often have redundant PSUs for resilience, so that one unit can carry the load if the other fails. Uninterruptible power supplies (UPS) and surge protectors add protection against outages and spikes, which are a concern in a server room or a computer lab.

What to monitor in the power supply

• PSU status lights: A steady power light means normal operation. A red, amber or flashing light often indicates a fault, so refer to the device manual.
• Mains voltage: Unstable or very low voltage damages equipment. A multimeter can be used to measure the output of a socket or UPS.
• UPS condition: Check the battery indicator, the load level and the time the UPS can supply power. Test the battery regularly according to the manufacturer's guidance.
• Surge protection: Check that surge protectors are in use and that their protection indicator is still active, since protection wears out with age.
• Redundant PSUs: Confirm that both units are working, as a failed second unit leaves the device without backup.
• Power over Ethernet budget: A PoE switch has a limited total power for connected devices such as access points and IP phones. Overloading the budget causes devices to switch off.
• Earthing and connections: Loose power cords or poor earthing cause faults and are a safety risk.

Procedure for a power supply check in a network room

1. Look at the PSU and UPS indicator lights on each device and note any warning colours.
2. Check that power cords are fully inserted and not damaged or strained.
3. Check the UPS display for battery state and load, and compare the load with its rated capacity.
4. Measure the mains output with a multimeter, taking care to follow electrical safety rules. Mains supply in Kenya is about 240 V.
5. Check surge protector indicators and replace any that show protection has failed.
6. Record the readings and report any device that is not powering correctly.
7. Schedule a repair or replacement of any faulty unit before it causes an outage.

3.1.3.4 Network optimization

Network optimization in this chapter means using monitoring information to tune the existing network so that it performs as well as it can. Monitoring shows where traffic is heavy, where links are failing and which settings or layouts waste capacity. Replacing hardware and cables is dealt with under the optimizing topic, so here the focus is on adjustments that a technician can make with the equipment already in place. Example: a computer lab switch shows one port at constant high utilisation because a single computer is running large downloads, and moving that computer to a less busy switch relieves the uplink.

Using monitoring to optimise a network

• Set a baseline: Record normal utilisation, error counts and response times when the network is healthy. Later readings are then compared with this baseline to show a change.
• Find congested links: Ports and uplinks that stay heavily loaded for long periods show where traffic needs to be rebalanced.
• Remove faulty links: Replacing a damaged patch cord or re-terminating a poor connector often restores full speed on a slow port.
• Tidy cabling and layout: Reorganising cable runs, shortening patch cords and separating data cables from power cables reduces errors and interference.
• Balance the load: Moving heavy users to different switches or access points spreads traffic more evenly.
• Segment traffic: Grouping devices into separate VLANs reduces unnecessary broadcast traffic on larger networks.
• Prioritise important traffic: Quality of service settings give voice or transaction traffic priority over less urgent downloads.
• Keep devices updated: Applying manufacturer firmware updates fixes known faults and can improve performance.
• Remove unused devices: Disabling unused ports and removing old equipment reduces clutter and security risk.

Procedure for a monitoring-led optimisation review

1. Collect current utilisation, error and device status readings for the main switches, routers and access points.
2. Compare the readings with the recorded baseline and list the ports or links that differ.
3. Check the cables and connectors on the problem links with a cable tester and visual inspection.
4. Correct physical faults first, such as loose or damaged cables and failing power supplies.
5. Adjust settings or move devices to balance the traffic where one link is overloaded.
6. Monitor the network again for a suitable period to confirm that the changes improved performance.
7. Record the changes, the results and any remaining problems, and report those that need hardware or cable upgrades.

Chapter Summary

This chapter focused on the essential practices involved in performing computer network maintenance. It began by introducing the concept of computer network monitoring and maintenance, highlighting its role in ensuring network reliability and performance. The discussion then moved to the physical tools used in network monitoring, such as cable testers, crimping tools, and stripping tools, which are vital for diagnosing and repairing physical network components. Attention was given to monitoring the status of physical networking devices, including ports and interfaces, cables and connections, power supplies, and overall network optimization strategies. The chapter also covered troubleshooting techniques for identifying and resolving network issues effectively. Finally, it explained the importance of optimizing the network through upgrading hardware devices and network cables to maintain efficient and stable network operations.