Network Troubleshooting and Maintenance
Network Troubleshooting and Maintenance tests structured fault isolation and routine upkeep. Expect questions on interface error counters, speed and duplex mismatch, spanning-tree loops, connectivity tools, neighbor discovery, configuration backup, and recovery procedures.
Structured methods and tools
Bottom-up troubleshooting starts at cabling and interfaces, top-down starts at applications, and divide-and-conquer tests the most likely middle layer first. ping tests reachability and round-trip time; traceroute reveals the path by using incrementing TTL values; Telnet or SSH tests TCP port 22 or 23 reachability. show interfaces reveals CRC, input, output, and late-collision counters. show interfaces status shows speed, duplex, VLAN, and connected state. show cdp neighbors and show lldp neighbors identify directly connected devices. show mac address-table and show ip arp confirm Layer 2 and Layer 3 mappings. Cable testers and TDR locate opens and shorts.
Recognizing common faults
CRC and FCS errors usually point to cabling, interference, or a bad port. Late collisions on a half-duplex link strongly suggest a duplex mismatch. Runts can indicate collisions or faulty cabling. A spanning-tree loop produces broadcast storms, MAC address flapping, and high CPU. An interface in err-disabled state needs the cause corrected before recovery.
Worked example
Users on one floor report slow, intermittent connectivity. Gi0/12 shows rising late collisions and FCS errors; show interfaces status reports 100 Mb/s half-duplex while the attached device is fixed at 100 Mb/s full-duplex. The reasoning: late collisions occur after the slot time and are characteristic of one side running full-duplex while the other runs half-duplex. The fix is to set both ends to auto-negotiate or hard-set both to the same speed and duplex; replacing the cable would not correct a duplex mismatch. After correction, clear the counters and confirm they stop increasing.
Maintenance and traps
Back up configurations and IOS images before changes. Know that ping success does not prove the whole path or application health, and traceroute can show asterisks for filtered hops. Do not assume every input error is a bad cable: check both ends, speed and duplex, and error type. Verify that the correct VLAN is active and allowed on the trunk. Document the baseline so later counters can be compared.