[Top] [Table Of Contents] [Prev] [Next] [Index]

Solving Hardware Problems        1


Use Table 1-1 to identify and diagnose common hardware problems.

Common Hardware Problems

Problem

Possible Cause

Possible Solution

The show S0 command shows two different port states on an asynchronous port.

Asynchronous ports were configured before cables and modems were attached.

Verify the port state as explained on page 1-2.

The show S0 or show W1 command shows a high number of line errors occurring on a synchronous or asynchronous port.

Use show S0 or show W1 to determine that line errors are present (see page 1-2), and follow the steps in "Diagnosing Line Errors" on page 1-4.

The Ethernet port is not communicating with the network.

Incorrect configuration on the Ethernet port.

See "Diagnosing Ethernet Port Problems" on page 1-7.

Unspecified Ethernet problems are occurring.

Incorrect DIP switch position.

See "DIP Switch Position" on page 1-8 for a list of settings.

Ethernet port and DIP switch configuration is correct, but the Ethernet is still not working.

Malfunctioning external hardware.

See "Diagnosing Faulty Ethernet Hardware" on page 1-9.

The network is slow.

Overloaded network.

See "Diagnosing an Overloaded Ethernet Network" on page 1-12.

Data arrives at its destination in a corrupted state.

Overloaded network.

See "Diagnosing an Overloaded Ethernet Network" on page 1-12.

All external Ethernet hardware is functioning correctly, but the Ethernet is still not working.

Loose Ethernet daughterboard.

See "Diagnosing Ethernet Daughterboard Problems" on page 1-14.

Diagnosing Synchronous and Asynchronous Port Problems

The PortMaster records statistics and keeps a count of errors detected on its synchronous and asynchronous ports. Use show commands to display the current error count and status on a port, and correct problems as instructed in the following sections.

Disable a synchronous hardwired port (see page 1-6), if necessary.

Displaying Port Errors and Status

The show W1 command displays status information for synchronous ports, and the show S0 command displays status information for asynchronous ports. Analyze the type and number of port errors to help diagnose port problems. For more information on show command output, see the PortMaster Command Line Reference.

Sample synchronous port output:

Command> show w1

 

 

----------------------- Current Status - Port W1 ---------------------

Status:

ESTABLISHED

 

 

Input:

915287284

Abort Errors:

56/1

Output:

3214289999

CRC Errors:

27

Pending:

0

Overrun Errors:

0

TX Errors:

0

Frame Errors:

15

Modem Status:

DCD+ CTS+

 

 

Sample asynchronous command output:

Command> show s10

 

 

----------------------- Current Status - Port S10 ---------------------

Status:

ESTABLISHED

 

 

Input:

1392900

Parity Errors:

0

Output:

453743

CRC Errors:

27

Pending:

0

Overrun Errors:

0

TX Errors:

0

Frame Errors:

0

Modem Status:

DCD+ CTS+

Abort Errors

0

Data integrity is directly related to the integrity of OSI Layers 1 (physical) and 2 (data link). High abort, CRC, and frame error rates result in a loss of data integrity and degraded service. Low data link integrity can cause serious connection difficulties.

Abort Errors (Synchronous Port Only)

An abort error occurs when a connection was not established on the synchronous port and the user is trying to connect again. Each time a connection attempt is unsuccessful, the error count increments by 1.

Abort errors are sometimes displayed as two counts separated by a slash:

Abort Errors: frame errors/ device errors

Frame Errors

Asynchronous frame errors occur when a frame is corrupted in transit-usually because of a hardware failure in a line or modem. A frame is considered invalid if it does not terminate with at least 1 stop bit. If a frame error occurs, the counter is incremented and the PortMaster automatically attempts to resynchronize by identifying the incorrect stop bit as the start bit for the next character. A new character is then constructed beginning with this new start bit.

Synchronous frame errors occur when there is a malfunctioning clock, for example, clock slippage. On synchronous ports, frame errors are sometimes displayed as two numbers separated by a slash:

Frame Errors: small packet errors/ large packets errors

Diagnosing Line Errors

Line errors encompass frame errors, abort errors, and CRC errors. Use the following steps if a large number of line errors appear in the show command output:

  1. Determine the rate of line errors.

    Are line errors incrementing at a constant rate? A constant increase in line errors indicates a hardware failure (cables, connectors, and so on). If the line errors happened all at once, they were probably caused by a single event. For example, unplugging a serial cable from the PortMaster results in a sudden, one-time burst of frame errors.

  2. Determine if the line errors are limited to one port.

    If you notice that a particular port has numerous line errors, replace the serial cable, device-modem or channel service unit/digital service unit (CSU/DSU), and/or telephone wire with that of another port. If the problem remains on the port, check the PortMaster. If the problem follows the line, begin replacing cables, modems, and so on, to isolate the source of the line errors.

  3. Verify the modem or CSU/DSU.

    Modems and CSU/DSUs can cause line errors if they are defective or misconfigured. You can use two bit error rate tester (BERT) sets to isolate the problem as well as determine the direction of the errors.

  4. Check the environment.

    Other devices cause noise that can result in line errors. Be sure to keep your data cables away from fluorescent lights, monitors, magnets, televisions, arc welders, air conditioners, vacuum cleaners, and other noise-producing devices. If possible, identify the communication media the telephone company is using (microwave, satellite, and so on). Different media are affected by weather conditions such as snow, lightning, and fog. These weather effects can include line errors.

  5. Check your serial cables.

    Ensure that the cable (and possibly the connectors and adapters) between the PortMaster and peripheral devices (such as a modem or a CSU/DSU) is shielded to minimize the noise from other serial cables and power connectors. If the errors are occurring on a synchronous port, verify that you are using Lucent V.35 cables. If the errors are occurring on an asynchronous port, verify that you are using flat ribbon cables-and not Category 5 unshielded twisted pair (UTP) cables. Ensure that cable lengths are within the standard industry recommendations.

  6. Try replacing the cable and/or adapters.

    Cables and adapters can be faulty. Verify cable, connector, and pin integrity visually, and ensure correct pinout specifications as specified in the hardware installation guide that was shipped with your PortMaster. When replacing cable hardware, change only one part at a time so you can more easily pinpoint the source of the problem. Use store-bought cables, if possible, to prevent human error. Be especially suspicious of DB-25-to-RJ-45 connectors because these often have problems. If you have a synchronous hardwired port, disable it before replacing cables. See "Disabling a Synchronous Hardwired Port" on page 1-6.

  7. Verify the telephone wire and punchdown block.

    Check the wiring in your building, in the punchdown block, and from the punchdown block to the modem or CSU/DSU. If wiring is faulty (for example, connections are loose or wires are crossed), or you have bare copper or ordinary telephone wiring running for a few feet, the PortMaster is more susceptible to line errors.

  8. Have your lines tested for noise.

    Telephone lines are susceptible to noise originating at the telephone company. This noise is often the cause of line errors. The telephone company might conduct a noise test for free. Noise can be intermittent, so make a note of when the line errors are occurring. Intermittent noise can be related to the environment, such as rain seeping into poorly sealed cables.
    Keep accurate records of line tests from the day of line commissioning to identify any changes or trends in line integrity that might require a call to the telephone company. If line errors occur during 0.5 and 1 percent of line use, notify the telephone company and request that they run line tests.

Verifying Port State for Old and New Cards

If you configure asynchronous ports before you attach cables and modems, you might see two different port states when you use the show S0 command. Ports on the main system card might show a status of IDLE, while ports on older expansion cards might show a status of USERNAME.

This behavior is normal. On older expansion cards, the value of carrier detect (CD) signal floats high in contrast to the carrier detect signal on the main system card. On more recent expansion cards, the carrier detect signal is pulled low as it is on the main system card.

On both old and new cards, a port should show a status of IDLE when modems are attached with hardware flow control set and modem control turned on for the port.

Disabling a Synchronous Hardwired Port

You might need to disable a synchronous hardwired port to diagnose problems with the port. For example, disable a hardwired port when you are replacing cables to diagnose line errors. (See "Diagnosing Line Errors" on page 1-4.)

To disable a synchronous hardwired port, use the following commands:

Command> set W1 protocol ppp
Protocol for port W1 changed from frame relay to ppp

Command> set W1 destination 0.0.0.0
Port W1 destination changed from 255.255.255.255 to 0.0.0.0

Command> reset W1
Resetting port W1

Command> save all

Diagnosing Ethernet Port Problems

When troubleshooting an Ethernet port, use the following procedure:

  1. Ensure that the Ethernet port has an IP address on the same IP network as the rest of your network by using the show ether0 command:

     

    Command> show ether0

     

    Ethernet Status:

    IP - Enabled

    IPX - Enabled

    Interface Addr:

    pm2.edu.com (192.168.96.6)

    Netmask:

    255.255.255.0

     

    Broadcast Address:

    192.168.96.0

     

    IPX Network:

    FEEDFEFE

     

    IPX Frame Type:

    ETHERNET_802.3

     

    Ethernet Address:

    00:c0:05:01:06:20

     

    Routing:

    Broadcast, Listen (On)

     

    Input Filter:

     

     

    Output Filter:

     

     

  2. Ensure that the Ethernet port is using the same netmask as the rest of the network.
Connect to another router on the same subnet and check that router's configuration.
  1. Verify the Ethernet address.
The Ethernet address shown in the show ether0 output displays the Ethernet hardware media access control (MAC) address. All PortMaster products have 00:c0:05 as the first three pairs of characters. If the address does not show these characters, or consist of all zeros, the PortMaster EPROM could be faulty. In addition, a faulty network interface card (NIC) on a PC has been known to cause the PortMaster to display all zeros.
  1. If necessary, set the IP address and netmask with the following commands:

    Command> set ether0 address Ipaddress

    Command> set ether0 netmask Ipmask

  2. If you are using IPX verify the following:

  3. If necessary, set the IPX network number and frame type with the following commands:

    Command> set ether0 ipxnet Ipxnetwork

    Command> set ether0 ipxframe ethernet_802.2|ethernet_802.2_ii|ethernet_802.3|ethernet_ii

  4. If necessary, re-enable IP or IPX, use the following commands:

    Command> set ether0 ip enabled

    Command> set ether0 ipx enabled

DIP Switch Position

If you are experiencing Ethernet problems, check the DIP switches to ensure that they are in the correct position:

DIP 4

DIP 5

Ethernet

Up

Up

10BaseT (twisted pair)

Up

Down

AUI (10Base5)

Down

Down

BNC (10Base2)

If the DIP switches are not in the correct position, reset them and turn the power off and on.

Note ¯ Office Router models do not have DIP switches 4 and 5. To specify the Ethernet type on an Office Router, set DIP switch 3 down to use AUI or up to use 10BaseT. BNC is not supported on Office Router models.

Try pinging another host on the network and checking the Address Resolution Protocol (ARP) requests to see if you have network connectivity. See "Verifying Network Connections" on page 3-9 for instructions. You can also try pinging the Ether0 interface to verify that it is active.

Diagnosing Faulty Ethernet Hardware

If you suspect faulty hardware is the cause of your Ethernet problem, use the show netstat command to display network statistics. If show netstat command output points to faulty hardware-cables, connectors, hub, or network interface card (NIC), for example-use the procedures in the following sections to help you isolate the problem. If possible, use Category 5 wire to assure a quality connection.

To use network statistics to determine Ethernet hardware health, do the following:

  1. Enter the following command:

    Command> show netstat

    Name

    Ipkts

    Ierrs

    Opkts

    Oerrs

    Collis

    Resets

    Queue

    ether0

    207757

    0

    215161

    0

    223

    0

    0

  2. Note the statistics:

These statistics can also indicate an overloaded Ethernet network. See "Diagnosing an Overloaded Ethernet Network" on page 1-12.
  1. If network statistics indicate faulty Ethernet hardware, use the procedures in the following sections to isolate and solve the problem.
In most of these procedures, you isolate the failure by replacing each possibly faulty component with one that you know is working. If you have a combination of Ethernet types, you must verify each type separately.
  1. If replacing Ethernet hardware does not solve the problem, try using another Ethernet type, if possible-for example, replace 10BaseT with BNC (10Base2).
The PortMaster might have a faulty Ethernet bus on its network interface card (NIC). Because AUI (10Base5) and BNC use a different Ethernet bus from 10BaseT, these types can function when 10BaseT does not, and vice versa.
If this behavior occurs, call Lucent Remote Access Technical Support.

Checking 10BaseT (RJ-45, Twisted Pair) Hardware

Use the following procedure to check 10BaseT Ethernet external hardware. After each step, ping another host on the network and check for ARP requests. (See "Verifying Network Connections" on page 3-9)

  1. Verify that you are using a Category 5 twisted pair cable with an RJ-48C DB modular connector.

  2. Verify that the link status LEDs on the PortMaster and the hub for the associated port are lit.

    If the LEDs are not lit, there might be a problem with link integrity. Verify the following:

  3. Replace the Ethernet cable with a working cable.

  4. Move the Ethernet cable to a different, working port on the Ethernet hub.

  5. Replace the hub with a working hub.

  6. Make sure that the cables are securely crimped into their connectors.

Checking BNC (10Base2, Coaxial, Thinnet) Hardware

Use the following procedure to check BNC Ethernet external hardware. After each step, ping another host on the network and check for ARP requests. (See "Verifying Network Connections" on page 3-9)

  1. Verify that you are using a RG-58 A/U 50-ohm coaxial cable, and that it is securely in place.

    BNC connectors often come undone.

  2. Verify that you are using T-connectors on all BNC Ethernet connections.

  3. Verify that 50-ohm terminator caps are installed on the ends of the Ethernet cable(s).

  4. Replace T-connectors with working T-connectors.

  5. Replace terminators with working terminators.

  6. Replace cables with working cables.

Checking AUI (10Base5, Ethernet D, Thicknet) Hardware

Use the following procedure to check AUI Ethernet external hardware. After each step, ping another host on the network and check for ARP requests. (See "Verifying Network Connections" on page 3-9.)

  1. Verify that you are using an RG-11 50-ohm coaxial cable with a DB-15 female connector.

  2. Verify that DIP switches are correctly set.

  3. Replace cables with working cables.

  4. Make sure that any signal quality editor (SQE) switches on AUI transceivers are turned off.

  5. Replace transceivers with working transceivers.

  6. Make sure connections to transceivers are secure.

Diagnosing an Overloaded Ethernet Network

To determine if an overloaded network is causing Ethernet problems, do the following:

  1. Enter the following command:

    Command> show netstat

    Name

    Ipkts

    Ierrs

    Opkts

    Oerrs

    Collis

    Resets

    Queue

    ether0

    207757

    0

    215161

    0

    223

    0

    0

  2. Check the collision (Collis) count and determine the collision rate.
Compute the collision rate (also known as the saturation rate) by using the following equation:
number of collisions (Collis)/number of output packets (Opkts)

A number of collisions greater than 5 percent of the total output (Opkts) indicates either an overloaded network or an Ethernet hardware failure-a problem with a cable, hub, or network interface card (NIC), for example.
 Table 1-2 helps you determine Ethernet condition from collision rates, errors, and resets.
Note ¯ The Ethernet traffic throughput expressed in Table 1-2 is a percentage of the total possible throughput (10Mbps). Your LAN might be able to handle a collision rate of 30 percent to 60 percent during short traffic bursts, but might not be able to sustain this rate.

Ethernet Conditions

show netstat Output

 

 

 

Ierrs

Oerrs

Resets

Collision Rate

Ethernet Traffic Throughput

Ethernet Condition

0

0

0

Less than 1%

0% to 10%

Excellent

0

0

0

1 to 5%

10% to 25%

Good

0

0

0

5 to 10%

Marginal

1 to 9

1 to 9

1 to 9

Above 10%

Above 25%

Poor

10+

10+

10+

Above 10%

 

Unsatisfactory

10+

10+

1 or more/hour

Above 30%

 

Inoperative

  1. If necessary, verify that Ethernet hardware is operational.
See "Diagnosing Faulty Ethernet Hardware" on page 1-9.
  1. If the network is overloaded, do one of the following to reduce traffic on the network:

Diagnosing Ethernet Daughterboard Problems

Note ¯ The following procedure applies to the PM-2, PM-25, and IRX PortMaster models only.

Once you have verified all your other Ethernet hardware (port, cables, hubs transceivers, T-connectors, terminators), inspect the Ethernet daughterboard inside the PortMaster chassis.

Warning ¯ Do not attempt this procedure if you are unfamiliar with servicing computer hardware.

To inspect the Ethernet daughterboard inside the PortMaster chassis, do the following:

  1. Turn off the PortMaster.

  2. Unplug the PortMaster power cord.

  3. Disconnect all cables from the PortMaster.

  4. Remove the single screw at the top center of the back panel of the PortMaster with a Phillips screwdriver.

  5. With both hands, gently push the cover forward to dislodge it from the PortMaster.

  6. If you are checking a PortMaster 3, remove the motherboard by removing the two screws.

  7. Visually inspect the Ethernet daughterboard mounted below the main board.

    Check to see that it is securely mounted and that the Ethernet daughterboard is connected to the main board.

  8. If the Ethernet daughterboard appears to be loose, reseat it.

  9. If you are checking a PortMaster 3, replace the motherboard and tighten the two screws.

  10. Reinstall the cover on the PortMaster and replace the screw in the back panel.

  11. Connect all required cables to the PortMaster.

  12. Plug in the power cord and turn the PortMaster on.


[Top] [Table Of Contents] [Prev] [Next] [Index]

spider@livingston.com
Copyright © 1998, Lucent Technologies. All rights reserved.