Fiber Internet Troubleshooting: From Optical Signal to IP Service

A structured troubleshooting path for LOS, low optical power, intermittent links, PPPoE failures, and internet problems over fiber access networks.

Category: Fiber NetworkingUpdated: September 17, 2026Reading time: 6 min
Before production changes: back up your configuration, verify interface names and IP ranges, and use a maintenance window for remote or customer-facing systems.

Fiber troubleshooting is fastest when you move from the physical layer upward. Start with power and continuity, then registration, VLAN/service mapping, authentication, IP addressing, routing, DNS, and finally application performance. Skipping layers often creates unnecessary configuration changes.

Video walkthrough

1. Check alarms and LEDs

A red LOS indicator usually means the ONT is not receiving a usable downstream optical signal. Check patch cords, connectors, splitters, closures, and feeder status before changing router configuration. If the PON light is stable but internet service fails, continue to higher layers.

2. Inspect connectors correctly

Dirty connectors are a common cause of excessive loss. Use proper fiber inspection and cleaning procedures. Do not look into an active fiber connector. Protect cleaned connectors with caps and avoid touching ferrule end faces.

3. Compare optical readings with the design

Read the OLT/ONT optical levels and compare them with the vendor's supported range and the expected budget for the route. A sudden change from the historical reading is often more useful than a single isolated measurement.

4. Confirm ONT authorization and service mapping

Verify that the correct serial number or LOID is registered on the intended PON port and profile. Then confirm VLANs, GEM/service mapping, bridge or router mode, and bandwidth profiles.

5. Test subscriber authentication

For PPPoE networks, check whether the client reaches the PPPoE server, whether credentials are accepted, and whether an IP address is assigned. A failed PPPoE discovery stage points to a different set of causes than an authentication rejection.

6. Verify IP, routing, and DNS separately

Once the subscriber receives an IP, ping the gateway, then a public IP, then resolve a domain. This three-step test separates local Layer 3 reachability, upstream internet reachability, and DNS resolution.

7. Investigate performance last

If connectivity is stable but speed or latency is poor, check interface errors, uplink utilization, queue limits, CPU, PON utilization, Wi-Fi conditions, and upstream congestion. Do not replace optical components solely because of a slow speed test when optical readings and error counters are healthy.

Frequently asked questions

What does LOS mean?

It generally indicates loss of optical signal at the ONT/ONU.

Can a dirty connector cause intermittent service?

Yes. Contamination can increase loss or reflections and create unstable optical conditions.

Why can PPPoE fail when the PON light is green?

Because optical registration can be healthy while VLAN, service mapping, credentials, or upstream configuration is wrong.

Next step

Document the result of your test and keep a known-good export. If your network behaves differently from the example, diagnose the topology rather than adding more rules blindly. Continue with a related guide from the tutorial library.