Catalyst 3850 Switch Hardware Installation Guide
Downloads: This chapterpdf (PDF - 1.15 MB) The complete bookPDF (PDF - 7.71 MB) | The complete bookePub (ePub - 3.07 MB) | Feedback



Diagnosing Problems

The LEDs on the front panel provide troubleshooting information about the switch. They show POST failures, port-connectivity problems, and overall switch performance. You can also get statistics from Device Manager, from the CLI, or from an SNMP workstation.

Switch POST Results


POST failures are usually fatal. Contact your Cisco technical support representative if your switch does not pass POST.

Switch LEDs

If you have physical access to the switch, look at the port LEDs for troubleshooting information about the switch. See the LED section for a description of the LED colors and their meanings.

Switch Connections

Bad or Damaged Cable

Always examine the cable for marginal damage or failure. A cable might be just good enough to connect at the physical layer, but it could corrupt packets as a result of subtle damage to the wiring or connectors. You can identify this situation because the port has many packet errors or the port constantly flaps (loses and regains link).

  • Examine or exchange the copper or fiber-optic cable with a known, good cable.

  • Look for broken or missing pins on cable connectors.

  • Rule out any bad patch panel connections or media convertors between the source and the destination. If possible, bypass the patch panel, or eliminate faulty media convertors (fiber-optic-to-copper).

  • Try the cable in another port or interface, if possible, to see if the problem follows the cable.

Ethernet and Fiber-Optic Cables

Make sure that you have the correct cable for the connection.

  • For Ethernet, use Category 3 copper cable for 10 Mb/s UTP connections. Use either Category 5, Category 5e, or Category 6 UTP for 10/100/1000 Mb/s connections.

  • For fiber-optic cables, verify that you have the correct cable for the distance and port type. Make sure that the connected device ports both match and use the same type encoding, optical frequency, and fiber type.

  • For copper connections, determine if a crossover cable was used when a straight-through was required or the reverse. Enable auto-MDIX on the switch, or replace the cable.

Link Status

Verify that both sides have link. A single broken wire or a shutdown port can cause one side to show link even though the other side does not have link.

A port LED that is on does not guarantee that the cable is fully functional. The cable might have encountered physical stress that causes it to function at a marginal level. If the port LED does not turn on:

  • Connect the cable from the switch to a known good device.

  • Make sure that both ends of the cable are connected to the correct ports.

  • Verify that both devices have power.

  • Verify that you are using the correct cable type.

  • Look for loose connections. Sometimes a cable appears to be seated, but is not. Disconnect the cable and then reconnect it.

10/100/1000 Port Connections

A port appears to malfunction:

  • Use the Mode button to show the status for all ports.

  • Use the show interfaces privileged EXEC command to see if the port is error-disabled, disabled, or shutdown. Reenable the port if necessary.

10/100/1000 PoE+ Port Connections

A powered device connected to a PoE port does not receive power:

  • Use the Mode button to show the PoE status for all ports.

  • Use the show interfaces privileged EXEC command to see if the port is in error-disabled, disabled, or shutdown. Reenable the port if necessary.

  • Verify that the power supply installed in the switch meets the power requirements of your connected devices.

  • Verify that there is sufficient PoE power budget to provide power to the attached device. Use the show power inline global configuration command to check on the available PoE power budget.

  • Verify the cable type. Many legacy powered devices, including older Cisco IP phones and access points that do not fully support IEEE 802.3af, might not support PoE when connected to the switch by a crossover cable. Replace the crossover cable with a straight-through cable.


    Noncompliant cabling or powered devices can cause a PoE port fault. Use only standard-compliant cabling to connect Cisco prestandard IP Phones and wireless access points or IEEE 802.3af-compliant devices. You must remove any cable or device that causes a PoE fault.

SFP and SFP+ Module

Use only Cisco SFP or SFP+ modules in the switch. Each Cisco module has an internal serial EEPROM that is encoded with security information. This encoding provides a way for Cisco to identify and validate that the module meets the requirements for the switch.

  • Inspect the SFP module. Exchange the suspect module with a known good module. Verify that the module is supported on this platform. (The switch release notes on list the SFP modules that the switch supports.)

  • Use the show interfaces privileged EXEC command to see if the port or module is error-disabled, disabled, or shutdown. Reenable the port if needed.

  • Make sure that all fiber-optic connections are properly cleaned and securely connected.

Interface Settings

Verify that the interface is not disabled or powered off. If an interface is manually shut down on either side of the link, it does not come up until you reenable the interface. Use the show interfaces privileged EXEC command to see if the interface is error-disabled, disabled, or shutdown on either side of the connection. If needed, reenable the interface.

Ping End Device

Ping from the directly connected switch first, and then work your way back port by port, interface by interface, trunk by trunk, until you find the source of the connectivity issue. Make sure that each switch can identify the end device MAC address in its Content-Addressable Memory (CAM) table.

Spanning Tree Loops

STP loops can cause serious performance issues that look like port or interface problems.

A unidirectional link can cause loops. It occurs when the traffic sent by the switch is received by its neighbor, but the traffic from the neighbor is not received by the switch. A broken fiber-optic cable, other cabling problems, or a port issue could cause this one-way communication.

You can enable UniDirectional Link Detection (UDLD) on the switch to help identify unidirectional link problems.

Switch Performance

Speed, Duplex, and Autonegotiation

If the port statistics show a large amount of alignment errors, frame check sequence (FCS), or late-collisions errors, this might mean a speed or duplex mismatch.

A common issue with speed and duplex occurs when duplex and speed settings are mismatched between two switches, between a switch and a router, or between the switch and a workstation or server. Mismatches can happen when manually setting the speed and duplex or from autonegotiation issues between the two devices.

To maximize switch performance and to ensure a link, follow one of these guidelines when changing the duplex or the speed settings.

  • Let both ports autonegotiate both speed and duplex.

  • Manually set the speed and duplex parameters for the interfaces on both ends of the connection.

  • If a remote device does not autonegotiate, use the same duplex settings on the two ports. The speed parameter adjusts itself even if the connected port does not autonegotiate.

Autonegotiation and Network Interface Cards

Problems sometimes occur between the switch and third-party network interface cards (NICs). By default, the switch ports and interfaces autonegotiate. Laptops or other devices are commonly set to autonegotiate, yet sometimes autonegotiation issues occur.

To troubleshoot autonegotiation problems, try manually setting both sides of the connection. If this does not solve the problem, there could be a problem with the firmware or software on your NIC. You can resolve this by upgrading the NIC driver to the latest version.

Cabling Distance

If the port statistics show excessive FCS, late-collision, or alignment errors, verify that the cable distance from the switch to the connected device meets the recommended guidelines.

Clearing the Switch IP Address and Configuration

If you have configured a new switch with a wrong IP address, or if all of the switch LEDs start blinking when you are trying to enter Express Setup mode, you can clear the configured IP address. The switch returns to the factory default settings.


This procedure clears the IP address and all configuration information stored on the switch. Do not follow this procedure unless you want to completely reconfigure the switch.

    Step 1   If you are using Cisco IOS XE Release 3E or later releases, enter the erase startup-config privileged EXEC command to clear the contents of your startup configuration.
    Step 2   Press and hold the Mode button. The switch LEDs blink after about 2 seconds. If the switch is not configured, the LEDs above the Mode button turn green. You can omit the next step.
    Step 3   Continue holding down the Mode button. The LEDs stop blinking after 8 seconds, and then the switch reboots.

    You can now configure the switch by using Express Setup as described in the switch getting started guide.

    You can also configure the switch by using the CLI setup procedure described in the appendix.

    Replacing a Failed Data Stack Member

      Step 1   The replacement switch must be a Catalyst 3850 switch.
      Step 2   Power down the failed switch. Remove AC or DC input power, and if the switch is part of a StackPower stack, remove the StackPower cables.
      Step 3   Make sure the replacement switch is powered off, and then connect it to the stack. If you had manually set the member numbers for the switch stack, manually assign the member number of the failed switch to the replacement switch. To manually assign the stack member number, see the switch software configuration guide on
      Step 4   Make the same Gigabit Ethernet connections on the replacement switch that were on the failed switch.
      Step 5   Reinstall any modules, and make cable connections.
      Step 6   Power on the replacement switch.

      The replacement switch has the same configuration for all the interfaces as the failed switch and functions the same as the failed switch.