Focusing on on-site troubleshooting steps, judgment criteria, command quick lookup, and emergency repair, it covers four core scenarios: hardware, optical path, business, and network management. Following the principle of "restore first, then locate," it can quickly handle situations within 10 minutes.
Ⅰ.Hardware failures (prioritize troubleshooting, essential for the foundational layer)
1.Power supply failure (entire machine offline/frequent restarts)
（1）Diagnosis (Practical Steps)
（1.1） Check indicator lights: No power to the entire unit → Power module light (PWR) is off; Single module failure → Redundant power supply light is red/off.
（1.2） Measure voltage: Use a multimeter to measure the AC input (220V±10%) and the device power output (-48V/12V, deviation exceeding ±5% is abnormal).
（1.3）Check alarms: Check the network management system/serial port for "Power Fault" and "Voltage Abnormal" alarms.
（2）Quick Repair
（2.1）Loose connection: Reconnect the power cord, tighten the circuit breaker/terminals, and replace the socket.
（2.2）Module failure: Switch to a redundant power module; if no redundancy is available, replace with the same model (power off before proceeding).
（2.3）Abnormal voltage: Temporarily connect a UPS; contact the power supply department for emergency repairs.
2.Fan/heat dissipation failure (high temperature alarm/single board crash)
（1）Diagnosis (Practical Steps)
（1.1）Check Status: The network administrator checks the fan speed (abnormal if below 50% of rated speed) and equipment temperature (alarm if above 60℃).
（1.2） Check Air Ducts: Inlet/outlet air vents are unobstructed; fan is quiet/not running.
（2）Quick Repair
（2.1）Blockage: Clean dust, remove obstructions, and boost the server room air conditioning (temporarily use industrial fans as auxiliary systems).
（2.2） Fan failure: Replace the faulty fan module and restart the fan control process.
3. Single board/optical module failure (port LOS/no data)
（1）Diagnosis (Practical Steps)
（1.1） Inspect the lights: The RUN light on the board is off/constantly red (fault); the LOS light on the PON/uplink port is on (no light received).
（1.2） Measure optical power: PON port received light at -8 to -24dBm (normal), <-28dBm is weak, >-6dBm is too strong; uplink port received light at -10 to -25dBm (normal).
（1.3）Cross-test: Connect the faulty port to a normal optical module/fiber; insert the faulty board into a spare slot.
（2）Quick Repair
（2.1）Poor contact: Reseat the board/optical module and clean the gold fingers (with anhydrous alcohol swabs).
（2.2） Module failure: Replace with a single-mode optical module of the same rate (Class B+ for PON ports, 10G/1G matching for uplink ports).
（2.3） Board failure: Restart the board (command/physical reset). If ineffective, replace with a spare board and import the configuration.
II. Optical path faults (high-frequency faults, segmented troubleshooting is the most effective)
1. PON port optical path (ONU batch offline/low light)
（1）Diagnosis (Practical Steps)
（1.1）Check the OLT side: PON port LOS light on → no light received; optical power meter readings below -28dBm indicate weak light.
（1.2） Segmented troubleshooting: OLT PON port → ODF pigtail → splitter → trunk fiber → branch fiber (measure optical power segment by segment).
（1.3）Check the splitter: loose ports, incorrect splitting ratio (1:32/1:64 mismatch), splitter malfunction.
（2）Quick Repair
（2.1）Pigtail Issues: Clean the connector (alcohol swabs + fiber cleaning paper), replace bent/damaged pigtails.
（2.2） Low Light: Repair fiber optic breaks, replace splitter ports, check splice loss (re-splitter if >0.5dB).
（2.3）Excessive Light: Add a 5~10dB optical attenuator to the splitter side to prevent module overload.
2. Uplink optical path (network-wide service interruption/packet loss)
（1）Diagnosis (Practical Steps)
（1.1）Check the port: If the uplink port's LOS/LOF is lit, the optical path is broken; measure the received optical power, below -25dBm is abnormal.
（1.2） Check the peer end: Check if the core switch/transmission equipment port is up, and if the link is configured correctly.
（1.3）Cross-test: Switch to a backup uplink port/fiber to determine whether the fault is on the OLT side or the transmission side.
（2）Quick Repair
（2.1） Switch to a backup uplink (prioritize service recovery) and temporarily enable redundant ports.
（2.2）Transmission failure: Contact transmission specialists for emergency repair and mark the faulty fiber segment.
III. Business Faults (User-side sensitive, troubleshoot by following the steps "Registration → Configuration → Authentication")
1. Broadband failure (unable to dial/691/678/slow internet speed)
（1）Diagnosis (Practical Steps)
（1.1） Check ONU: The network administrator checks if the ONU is online, and whether the LOID/SN matches. No registration → authentication failure.
（1.2）Check VLAN: Check if the OLT/ONU service VLANs are consistent, and whether the outer/inner VLANs are transparent.
（1.3）Check authentication: PPPoE dialing error 691 → account locked/incorrect password; 678 → link down/authentication server malfunction.
（1.4）Check bandwidth: PON port/uplink port bandwidth utilization exceeding 80% → congestion; check for incorrect rate limiting configuration.
（2）Quick Repair
（2.1）Registration failed: Reconfigure ONU LOID, restart ONU, and verify the OLT authentication template.
（2.2） Error 691: Unlock the account, reset the password, and check the Radius server connectivity.
（2.3）Error 678: Correct the VLAN configuration, establish a connection between the OLT and the core network, and restart the PPPoE process.
（2.3）Slow network speed: Adjust the speed limiting policy, isolate the faulty ONU (unauthorized router connection/attack), and increase uplink bandwidth.
2. Voice malfunction (no dial tone/noise/drop-off)
（1）Diagnosis (Practical Steps)
（1.1）Check Registration: ONU voice port SIP registration status (Registered). Not registered indicates incorrect server parameters.
（1.2）Check Link: Is the voice VLAN transparent? Is there packet loss when pinging the SIP server?
（1.3）Check QoS: Is the voice stream priority the highest (COS 6/7)? Is bandwidth being preempted by data services?
（2）Quick Repair
（2.1）Registration failed: Reconfigure SIP IP/port/account, and restart the ONU voice module.
（2.2）Noise/Call Dropouts: Optimize QoS, ensure voice bandwidth (≥100kbps/line), and investigate electromagnetic interference.
3.IPTV malfunction (stuttering/screen flickering/multicast failure)
（1）Diagnosis (Practical Steps)
（1.1） Check multicast: Is IGMP Snooping enabled on the OLT? Is the multicast VLAN configured correctly?
（1.2）Check bandwidth: Single-channel bandwidth ≥ 4Mbps; is the uplink multicast stream being transparent?
（1.3）Check ONU: Is the IPTV port bound to the multicast VLAN? Are the IGMP versions compatible (v2/v3)?
（2）Quick Repair
（2.1）Enable/restart IGMP function and correct multicast VLAN configuration.
（2.2） Lag: Increase multicast bandwidth and isolate users illegally occupying multicast resources.
IV. Network management/software failures (control layer, remote/local dual troubleshooting)
1. Network administrator offline (unable to be remotely managed)
（1）Diagnosis (Practical Steps)
（1.1） Check connectivity: Ping the OLT management IP; failure indicates a link/IP configuration error.
（1.2）Check SNMP: Verify the community word (read/write) and port (161) are correct, and check for firewall blocking.
（1.3）Check version: Ensure the OLT and network management versions are compatible (incompatibility will cause offline issues).
（2）Quick Repair
（2.1） Corrected management IP/SNMP configuration to establish a link between the network management server and the OLT.
（2.2）Version incompatibility: Temporarily use local serial port management; upgrade the OLT/network management version later.
2. Configuration error (service interruption/configuration loss)
（1）Diagnosis (Practical Steps)
（1.1）Check backup: Compare the current configuration with the most recent backup (differences indicate points of failure).
（1.2）Check logs: Examine error-level logs to pinpoint configuration errors or software bugs.
（2）Quick Repair
（2.1）Configuration loss: Import the most recent valid backup and restart the service board.
（2.2）Software bug: Upgrade the OLT firmware to a stable version or revert to a normal version.
V. General 10-minute rapid troubleshooting process (essential for on-site use)
1. Monitor alarms: Prioritize critical/major alarms (power supply, LOS, board failures), filter minor alarms.
2. Check hardware: Check indicator lights → measure power supply/optical power → plug/replace modules/boards.
3. Segmented troubleshooting: OLT → splitter → ONU (downlink); uplink port → transmission → core network (uplink).
4. Cross-testing: Connect faulty ports to known good components, and install faulty boards in spare slots.
5. Check services: Check ONU registration → VLAN → authentication → bandwidth in sequence.
6.Emergency recovery: First switch to a backup link/module, then investigate the root cause to minimize downtime.
VI. Quick Reference for Mainstream Brand Commands (Can be directly copied and executed)
|
Fault Scenario |
HuaWei（MA5600/5800） |
ZTE（C300/C600） |
FiberHome（AN5516） |
|
Check ONU Status |
display ont info 0/0/1 all |
show ont registration 0/0 |
display ont info 0/0/1 |
|
Check Optical Power |
display port info 0/0/1 |
show pon port optical-info 0/0/1 |
display optical-info 0/0/1 |
|
Check VLAN Configuration |
display vlan 100 |
show vlan 100 |
display vlan 100 |
|
Check Board Status |
display board 0/1 |
show card status |
display board 0/1 |
|
Backup Configuration |
save |
write |
save config |
|
Restart Single Board |
reset board 0/1 |
reboot card 0/1 |
reset board 0/1 |
VII. Precautions for Practical Operation
1. Safety First: Disconnect power/prevent static electricity before operation; wear anti-static wrist straps when plugging and unplugging circuit boards.
2. Spare Parts on Hand: Power modules, optical modules, fiber optic pigtails, and spare circuit boards are readily available for on-site replacement.
3. Configuration Backup: Back up configurations immediately after each modification, retaining three historical versions.
4. Log Retention: Export logs after a failure for subsequent root cause analysis.
Post time: Feb-11-2026
