Fast Link Pulse is associated with Ethernet auto-negotiation on 10/100 Mbps copper links. A problem with FLP or auto-negotiation can leave a port connected while causing poor throughput, duplex mismatch, FCS errors, late collisions, or intermittent communication. This guide explains how to troubleshoot these conditions on Cisco Catalyst switches and connected NICs.
What Fast Link Pulse Means on Ethernet
Fast Link Pulse is associated with the auto-negotiation mechanism used by 10/100 Mbps Ethernet.
During negotiation, link partners exchange information such as supported speed, half- or full-duplex capability, and other link parameters. If the expected negotiation information is not detected, the port may fall back to a lower operating mode.
| Interface type | FLP relevance |
|---|---|
| 10BASE-T and 100BASE-TX copper | Main FLP and auto-negotiation use case |
| 1000BASE-T copper | Uses Gigabit auto-negotiation and additional link behavior |
| 1000BASE-X fiber | Uses a different negotiation and link mechanism |
| 10G SFP+ and faster interfaces | Do not apply 10/100 FLP assumptions directly |
Fast Link Pulse should not be treated as a universal term for every Ethernet interface. The correct troubleshooting method depends on the interface type, peer device, cable, and actual operating speed.
Cisco’s Catalyst and NIC compatibility troubleshooting guidance describes FLP as part of the 10/100 Mbps auto-negotiation process.
How Cisco Auto-Negotiation Determines Speed and Duplex
When both link partners use auto-negotiation correctly, the devices select a compatible operating mode. The exact result depends on the capabilities advertised by both sides.
| Switch side | NIC or peer side | Typical result |
|---|---|---|
| Auto | Auto | Normal negotiation when both devices support it |
| 100 Mbps full-duplex | Auto | Possible duplex mismatch |
| Auto | 100 Mbps full-duplex | Possible duplex mismatch |
| 100 Mbps full-duplex | 100 Mbps full-duplex | Correct manual configuration if both sides match |
| Auto | 10 Mbps half-duplex | May fall back to 10 Mbps half-duplex |
| 10 Mbps half-duplex | 100 Mbps half-duplex | No link because of speed mismatch |
| Auto | Device does not send FLP | May fall back to 10 Mbps half-duplex |
A link that is up is not necessarily a healthy link. If one side is configured manually and the other side is left on Auto, the link may come up with different duplex assumptions.
- Reduced throughput
- Late collisions
- Frame check sequence errors
- Retransmissions
- Intermittent connectivity
- High interface error counters
Common Symptoms of FLP and Auto-Negotiation Problems
The port is connected at 10 Mbps
- The peer does not send FLP.
- The peer is fixed at 10 Mbps.
- Auto-negotiation is disabled on one side.
- The cable or media converter does not support the expected speed.
- A third-party NIC has a compatibility problem.
- The interface or driver is malfunctioning.
The port is connected at 100 Mbps but performs poorly
- Duplex mismatch
- FCS or CRC errors
- Late collisions
- One-sided manual configuration
- NIC driver problems
- Damaged cable or patch-panel connections
The link repeatedly goes up and down
Possible causes include physical-layer errors, poor cable quality, NIC driver or firmware problems, repeated negotiation attempts, media-converter issues, incorrect speed settings, or hardware problems on the connected device.
Both sides are set to Auto but the link still fails
- Replace the patch cable.
- Move the connection to another switch port.
- Bypass the patch panel.
- Test with a known-good device.
- Check the NIC driver.
- Inspect interface counters.
- Confirm that the media converter supports auto-negotiation.
- Consider matching manual settings on both ends only after the physical path is verified.
Cisco Catalyst Troubleshooting Workflow
The following workflow applies to copper access ports on Catalyst switches such as the C9200-48P-E and C9300-48P-E. Exact interface names and available commands depend on the platform and software release.
Check the actual interface state
Check the operational speed, duplex mode, input errors, CRC or FCS errors, late collisions, interface resets, and link-flap history.
Do not rely only on the administrative configuration. The operational speed and duplex values show what the link is actually using.
Check both ends of the link
The Cisco switch is only one side of the negotiation. Also inspect the NIC speed and duplex settings, operating system network configuration, server BIOS or NIC utility, remote switch interface, media converter settings, cable type, and patch-panel connections.
A switch set to Auto cannot correct a peer that is incorrectly forced to a different speed or duplex mode.
Test the physical path
- Replace the patch cable with a known-good cable.
- Connect the device to another switch port.
- Bypass intermediate patch panels.
- Connect a known-good endpoint.
- Compare the interface counters.
- Test the original device on another known-good switch.
If the problem follows the endpoint, investigate the NIC, driver, or endpoint configuration. If it stays with the port or cable path, investigate the switch interface, patching, or physical media.
Speed and Duplex Configuration Rules
Auto-negotiation is normally preferred when both devices support it. If manual configuration is necessary, configure both ends consistently.
The remote device must also be configured for 100 Mbps full-duplex. Do not configure one side as 100 Mbps full-duplex while leaving the other side on Auto unless the device documentation explicitly supports that combination.
Manual settings should be considered only after checking the cable, peer-device capability, NIC driver and firmware, media converters, interface counters, and current operating mode on both ends.
Disabling auto-negotiation can hide link drops and physical-layer problems. It should not be used as a general solution for switch-to-switch links.
Cisco Catalyst 1300 Considerations
The Cisco Catalyst 1300 Series uses a different management workflow from IOS XE-based Catalyst 9200 and 9300 switches.
For C1300, review the relevant Web UI or release-specific CLI settings for interface speed, duplex mode, auto-negotiation, port status, cable diagnostics, and interface counters.
The exact menu names and available options may vary by firmware version. Do not copy a complete IOS XE interface configuration directly to a C1300.
C1300 troubleshooting workflow
- Check the actual negotiated speed.
- Confirm the peer configuration.
- Test the cable and switch port.
- Check errors and link flaps.
- Verify the NIC or media converter.
- Apply matching manual settings only when required.
Why a Link Can Be Up but Still Perform Poorly
A link indicator only confirms that the physical layer has detected a usable signal. It does not prove that both sides agree on speed and duplex.
For example, the switch may operate at 100 Mbps full-duplex while the NIC falls back to 100 Mbps half-duplex. The interface remains up, but throughput becomes poor and collisions or retransmissions increase.
Check the following counters and logs:
- FCS errors
- Input errors
- Late collisions
- Runts
- Interface resets
- Repeated link transitions
If errors continue after replacing the cable and testing another port, update the NIC driver or firmware and compare the endpoint against a known-good device.
| Symptom | Likely cause | First check |
|---|---|---|
| Port falls back to 10 Mbps | Missing FLP or incompatible peer | Speed and duplex on both ends |
| Link is up but slow | Duplex mismatch | Operational duplex and late collisions |
| Port remains down | Speed mismatch or physical issue | Cable, peer speed, and interface state |
| FCS errors increase | Duplex mismatch, cable, or NIC problem | Error counters and physical path |
| Link repeatedly flaps | Cable, NIC, media converter, or port issue | Logs, counters, and cable replacement |
| Only one device has the problem | Endpoint-specific compatibility issue | NIC driver and operating system settings |
Frequently Asked Questions
Final Takeaway
Fast Link Pulse is one part of 10/100 Mbps Ethernet auto-negotiation, but a Cisco port that connects at the wrong speed usually requires a broader check of the peer device, cable, NIC, media converter, and interface counters.
Start with the actual operational state. Confirm that both sides use compatible settings, replace the physical path with known-good components, and check for duplex and FCS errors. Only force speed and duplex when both ends can be configured consistently and the reason for doing so is understood.
If the link remains unstable after these checks, send your Catalyst model, interface output, peer device, and cable details for review.
Send the Catalyst model, interface output, peer device, and cable details.