Troubleshooting

The LEDs on the front panel provide troubleshooting information about the switch. They show failures in the power-on self-test (POST), port-connectivity problems, and overall switch performance. For a full description of the switch LEDs, see the "LEDs" section.

You can also get statistics from the browser interface, from the command-line interface (CLI), or from a Simple Network Management Protocol (SNMP) workstation. Refer to the Cisco IOS Desktop Switching Software Configuration Guide, and the Cisco IOS Desktop Switching Command Reference on Cisco.com, or the documentation that came with your SNMP application for details.

This chapter describes these topics for troubleshooting problems:

Understanding POST Results

Table 3-1 lists the ten POST tests and their associated LEDs.

POST tests run automatically each time the switch is powered on. When the switch begins POST, the port LEDs turn green. The System LED flashes green, and the RPS LED turns off. As each test runs, the port LEDs, starting with number 1, turn off. The remaining port LEDs each turn off in sequence as the system completes a test.

When POST completes successfully, the port LEDs return to the status mode display, showing that the switch is operational. If a test fails, the port LED associated with the test turns amber, and the system LED turns amber.

Note POST failures are usually fatal. Call Cisco Systems if your switch does not pass POST.

Table 3-1: POST Test Descriptions

Switch LED Component Tested

LED 1

DRAM

LED 2

Flash memory

LED 3

CPU interface buffer

LED 4

CPU interface ASIC

LED 5

Switch core ASIC

LED 6

TCAM

LED 7

Ethernet controller ASIC

LED 8

Ethernet interfaces

LED 9

CPU interface buffer second stage

LED 10¹

Inline Power Controller Test

¹This POST test is available only on the Catalyst 3550-24PWR switches.

Diagnosing Problems

Common switch problems fall into these categories:

Poor performance
No connectivity
Corrupted software

Table 3-2 describes how to detect and resolve these problems.

Table 3-2: Common Problems and Their Solutions

Symptom Possible Cause Resolution

Poor performance or excessive errors
Duplex autonegotiation mismatch.

Refer to the Cisco IOS Desktop Switching Software Configuration Guide for information on identifying autonegotiation mismatches.

Cabling distance exceeded
Port statistics show excessive frame check sequence (FCS), late-collision, or alignment errors.

For 10/100 and 10/100/1000BASE-T connections:

The distance between the port and the attached device exceeds 100 meters.

If the switch is attached to a repeater, the total distance between the two end stations exceeds the cabling guidelines.

For GBIC port connections:

The distance between the GBIC port and the attached device exceeds the GBIC cabling guidelines.

Refer to the Cisco IOS Desktop Switching Software Configuration Guide for information on displaying port statistics.

Reduce the cable length to within the recommended distances.

Refer to your repeater documentation for cabling guidelines.

Refer to your GBIC documentation for cabling guidelines.

Bad adapter in attached device
Excessive errors found in port statistics.

STP checking for possible

Run adapter card diagnostic utility.

Wait 30 seconds for the port LED to turn green.

No connectivity Incorrect or bad cable
These are results of no link at both ends:

A crossover cable was used when a straight-through was required, or vice-versa.

The cable is wired incorrectly.

A crossover or straight-through cable is wired incorrectly.

STP checking for possible loops.

For the correct pinouts and the proper application of crossover vs. straight-through cables, see the "Two Twisted-Pair Cable Pinouts" section.

Replace with a tested good cable.

For 1000BASE-T connections, be sure to use a twisted four-pair, Category 5 cable.

Wait 30 seconds for the port LED to turn green.

Switch not recognizing a GBIC module.

Refer to the your GBIC documentation for more information.

Unreadable characters on the management console
Incorrect baud rate.

Reset the emulation software to 9600 baud.

Amber System LED
Nonfatal or fatal POST error detected.

Check the port LEDs as described in Table 3-1 to see which POST test failed.

Switch placed in error-disabled state after GBIC is inserted
Bad or non-Cisco-approved GBIC.

Remove the GBIC module from the switch, and replace it with a Cisco-approved module. Use the errdisable recovery cause gbic-invalid global configuration command to verify the port status, and enter a time interval to recover from the error-disable state.

Refer to the Cisco IOS Desktop Switching Command Reference for information on the errdisable recovery command.

System LED is amber
Internal fan fault detected.

Switch is overheating.

Nonfatal or fatal POST error detected.

Either check the switch itself or use the show env command to check for a failed switch fan.
If necessary, replace the switch at your convenience.

Use the show env command to check for an overtemperature condition. If one exists:

Place the switch in an environment that is between 32 and 113°F (0 and 45°C).

Make sure fan intake and exhaust areas are clear.

Use the show POST command to see which POST test failed.

Switch LED	Component Tested
LED 1	DRAM
LED 2	Flash memory
LED 3	CPU interface buffer
LED 4	CPU interface ASIC
LED 5	Switch core ASIC
LED 6	TCAM
LED 7	Ethernet controller ASIC
LED 8	Ethernet interfaces
LED 9	CPU interface buffer second stage
LED 10¹	Inline Power Controller Test

Symptom	Possible Cause	Resolution
Poor performance or excessive errors	Duplex autonegotiation mismatch.	Refer to the Cisco IOS Desktop Switching Software Configuration Guide for information on identifying autonegotiation mismatches.
	Cabling distance exceeded Port statistics show excessive frame check sequence (FCS), late-collision, or alignment errors. For 10/100 and 10/100/1000BASE-T connections: The distance between the port and the attached device exceeds 100 meters. If the switch is attached to a repeater, the total distance between the two end stations exceeds the cabling guidelines. For GBIC port connections: The distance between the GBIC port and the attached device exceeds the GBIC cabling guidelines.	Refer to the Cisco IOS Desktop Switching Software Configuration Guide for information on displaying port statistics. Reduce the cable length to within the recommended distances. Refer to your repeater documentation for cabling guidelines. Refer to your GBIC documentation for cabling guidelines.
	Bad adapter in attached device Excessive errors found in port statistics. STP checking for possible	Run adapter card diagnostic utility. Wait 30 seconds for the port LED to turn green.
No connectivity	Incorrect or bad cable These are results of no link at both ends: A crossover cable was used when a straight-through was required, or vice-versa. The cable is wired incorrectly. A crossover or straight-through cable is wired incorrectly. STP checking for possible loops.	For the correct pinouts and the proper application of crossover vs. straight-through cables, see the "Two Twisted-Pair Cable Pinouts" section. Replace with a tested good cable. For 1000BASE-T connections, be sure to use a twisted four-pair, Category 5 cable. Wait 30 seconds for the port LED to turn green.
No connectivity	Switch not recognizing a GBIC module.	Refer to the your GBIC documentation for more information.
Unreadable characters on the management console	Incorrect baud rate.	Reset the emulation software to 9600 baud.
Amber System LED	Nonfatal or fatal POST error detected.	Check the port LEDs as described in Table 3-1 to see which POST test failed.
Switch placed in error-disabled state after GBIC is inserted	Bad or non-Cisco-approved GBIC.	Remove the GBIC module from the switch, and replace it with a Cisco-approved module. Use the errdisable recovery cause gbic-invalid global configuration command to verify the port status, and enter a time interval to recover from the error-disable state. Refer to the Cisco IOS Desktop Switching Command Reference for information on the errdisable recovery command.
System LED is amber	Internal fan fault detected. Switch is overheating. Nonfatal or fatal POST error detected.	Either check the switch itself or use the show env command to check for a failed switch fan. If necessary, replace the switch at your convenience. Use the show env command to check for an overtemperature condition. If one exists: Place the switch in an environment that is between 32 and 113°F (0 and 45°C). Make sure fan intake and exhaust areas are clear. Use the show POST command to see which POST test failed.

Table of Contents

Troubleshooting

Understanding POST Results

Diagnosing Problems