# How to Run a Traceroute for Fibre Troubleshooting

Source: https://urbanx.co.za/knowledge-hub/fno-troubleshooting/run-traceroute-fibre-troubleshooting

**Answer:** Traceroute shows the path your data takes from your device to a destination. It helps identify where latency or packet loss occurs. If PON is solid and LOS is off, traceroute can determine whether the issue is internal, access-layer, or upstream network related.

Part of [FNO Troubleshooting & Fibre Diagnostics](https://urbanx.co.za/knowledge-hub/fno-troubleshooting) · 6 min read · by UrbanX Network Operations

## What the Problem Means

If your fibre is connected but you experience:

- High latency
- Packet loss
- Intermittent disconnections
- Service-specific issues

Traceroute helps isolate whether the fault is:

- Inside your home network
- At the access layer
- Within upstream routing

> **INFO:** Traceroute does not test fibre light signal. It tests packet path progression.

If your ONT shows:

- PON solid
- LOS off

Then the optical signal is stable, and traceroute can help identify the next layer.

## What Traceroute Actually Does

Traceroute maps each "hop" between:

- Your device
- Your router
- Your ISP gateway
- Upstream transit
- Final destination

Each hop reports:

- Response time (ms)
- Packet response
- Timeouts (if any)

Consistent packet loss or high latency at a specific hop may indicate a fault at that layer.

## How to Run Traceroute (Windows & macOS)

**Windows**

1. Press **Windows Key + R**
1. Type `cmd`
1. Enter:

   ```
   tracert 8.8.8.8
   ```
1. Press **Enter**

**macOS**

1. Open **Terminal**
1. Enter:

   ```
   traceroute 8.8.8.8
   ```
1. Press **Enter**

Replace `8.8.8.8` with the affected service IP if known.

## How to Interpret Traceroute Results

| Pattern | Meaning | Likely Layer |
| --- | --- | --- |
| High latency at Hop 1 | Internal network issue | Router / LAN |
| High latency from Hop 2 onward | ISP gateway issue | Access Layer |
| Packet loss only at final hop | Server filtering | External |
| Timeouts mid-route but traffic continues | ICMP filtering | Not always a fault |
| Consistent packet loss across all hops | Network congestion | Upstream |

> **WARNING:** Not all timeouts indicate failure. Some routers deprioritize ICMP responses.

## Step-by-Step Diagnostic Flow

1. **Confirm ONT Lights**
   - PON solid, LOS off.
2. **Test via Ethernet**
   - Avoid WiFi interference.
3. **Run Traceroute to Stable IP**
   - Example: 8.8.8.8
4. **Run Traceroute to Affected Service**
   - Compare results.
5. **Look for First Hop With Persistent Packet Loss**
   - That hop often indicates fault layer.
6. **Repeat During Peak and Off-Peak**
   - Identify congestion patterns.

If packet loss begins beyond your ISP gateway, the issue may be upstream routing.

## Definition: Hop and ICMP

- **Hop** — A router or gateway that forwards your packet along the path.
- **ICMP** — Internet Control Message Protocol — used by traceroute to measure response time from each hop.
- **ICMP Filtering** — May cause timeouts that do not represent real packet loss.

## South African Context

In South Africa:

- Routing often passes through regional aggregation points
- Peak-time congestion may affect mid-route hops
- Undersea cable faults affect international routes only

If traceroute shows stable first hops but latency spikes internationally, the issue is not your fibre line.

## When to Escalate

Escalate when:

- Packet loss starts at ISP gateway hop
- Multiple traceroutes show consistent mid-route packet drop
- Issue persists outside peak hours
- Wired testing confirms instability

Do not escalate if:

- Only final hop shows packet filtering
- WiFi was used during testing

Provide traceroute output when logging a ticket.

## When This Is NOT the Issue

If all of the following are true, this is not your fault domain:
- LOS is red
- PON is flashing

Then traceroute is not required. Signal must be restored first.

## Frequently Asked Questions

**Does traceroute show fibre light problems?**

No. It tests packet path, not optical signal.

**What IP should I test?**

Start with a stable public IP, then test affected service.

**Are timeouts always bad?**

No. Some routers block ICMP responses.

**Should I run traceroute on WiFi?**

No. Use Ethernet for accurate results.

**How many tests should I run?**

At least two — one during peak and one off-peak.