Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Verifying Link and Transceivers using Pseudo Random Binary Sequence (PRBS) Test

Verifying Link and Transceivers using Pseudo Random Binary Sequence (PRBS) Test

The Pseudo Random Binary Sequence (PRBS) test is a standard feature to verify link quality and transceiver operation. There are two possible diagnostic scenarios:

  • Bidirectional verification using remote loopback. Test sequence is initiated from the local end. Remote end is configured with the loopback and analysis of the test pattern is performed at local end as well.

  • Unidirectional. Test sequence is initiated from the local end. Test pattern is analyzed by the remote end.

In the first case, verification is bidirectional and loopback support is required on the remote end.

The following table lists the entity that enables the PRBS test on various MICs:

MIC Type

PRBS Test Engine

Remarks

JNP10003-LC2103

Packet Forwarding Engine (PFE)

-

JNP-MIC1

Packet Forwarding Engine (PFE)

-

JNP-MIC1-MACSEC

External physical layer (PHY) device

External physical layer (PHY) device is not capable of passing the PRBS pattern originating from the packet forwarding engine. Hence, the PRBS test is enabled on external physical layer (PHY) device.

The following table mentions the PRBS test details supported on various interfaces:

Interface Type

Interface Name

Interface Lane Characteristics

Remarks

10-Gigabit ethernet interface

“xe”

The interface is supported with 1 lane of 10Gbps speed

The PRBS test is executed on each lane supported. Hence, the show interfaces prbs-stats displays data for one lane.

40-Gigabit ethernet interface

“et”

The interface is supported with 4 lanes of 10Gbps speed.

The PRBS test is executed on each lane supported. Hence, the show interfaces prbs-stats displays data for four lanes.

100-Gigabit ethernet interface

“et”

The interface is supported with 4 lanes of 25Gbps speed.

The PRBS test is executed on each lane supported. Hence, the show interfaces prbs-stats displays data for four lanes.

Guidelines to perform Unidirectional Diagnostics

Enable the transmission of pseudo-random binary sequence pattern on one end of the link (TX), and enable monitoring on the other end (RX).

The PRBS statistics report on the receiving end reflects link quality.

Following are the steps to collect and view the PRBS statistics:

  1. Start a TX (direction 0) by issuing the following command:

    After executing the command, you can check the link status by executing show interfaces terse et-0/1/*:

  2. Start a RX (direction 1) by issuing the following command:

    After executing the command, you can check the link status by executing show interfaces terse et-1/1/4:

  3. After starting the statistics collection, you can view the collected statistics at RX by issuing the following command:

    For example:

    Checking PRBS statistics at RX:

    The PRBS test is successful, if the state is pass with error count 0.

  4. Stop the PRBS statistics collection by issuing the following command:

    For example:

    To stop the PRBS at RX:

    Verify the statistics at TX by executing the following command:

    To stop the PRBS at TX:

    After executing the command, you can check the link status at RX by executing show interfaces terse et-1/1/4:

    Check the link status at TX by executing show interfaces terse et-0/1/2:

    This command only disables the statistics collection and does not clear the statistics collected. To clear the collected statistics, issue the clear interfaces statistics command.

Guidelines to perform Bidirectional Diagnostics using Remote Loopback

Configure loopback on the remote end of the link. Test pattern generation and analysis is performed on the local end.

The PRBS statistics report on the receiving end reflects link quality.

Following are the steps to collect and view the PRBS statistics:

  1. Enable remote loopback.
    Note:

    You must disable Forward Error Correction (FEC), if loopback is configured in the router with JNP-MIC1 MIC at the remote end.

  2. Start a TX (direction 0) by issuing the following command:

    After executing the command, you can check the link status by executing show interfaces terse et-0/1/*:

  3. Start a RX (direction 1) by issuing the following command on the same host.
    Note:

    There is a change in the direction as 1.

  4. After starting the statistics collection, you can view the collected statistics at RX by issuing the following command:

    For example:

    Checking PRBS statistics at RX:

    The PRBS test is successful, if the state is pass with error count 0.

  5. Stop the PRBS statistics collection by issuing the following command:

    For example:

    To stop the PRBS at RX:

    Verify the statistics at TX by executing the following command:

    To stop the PRBS at TX:

    After executing the command, you can check the link status at RX by executing show interfaces terse et-0/1/2:

    This command only disables the statistics collection and does not clear the statistics collected. To clear the collected statistics, issue the clear interfaces statistics command.

Interface Card Specific differences

While collecting statistics, the JNP-MIC1-MACSEC and JNP-MIC1 MICs behave differently:

  • On JNP-MIC1-MACSEC MIC, if RX is not latching to any PRBS signal, then the "state" in the show interfaces interface-name prbs-stats displays as “Disabled" with Error count as 0, where the JNP-MIC1 MIC displays as failed with MAX error count.

    For Example: On JNP-MIC1-MACSEC MIC

    For Example: On JNP-MIC1 MIC

  • If any mismatch is encountered between the pattern-type and flip between TX and RX, the maximum error counts are observed in case of JNP-MIC1 MIC and state “disabled” in case of JNP-MIC1-MACSEC MIC (flip is only supported in JNP-MIC1-MACSEC).

  • In case of JNP-MIC1 MIC, if TX is interrupted, the RX displays the state as failed with error counts. Even if TX is started again, RX also must to be restarted to work properly. In case of JNP-MIC1-MACSEC MIC, if TX is interrupted, the RX displays the state as "Disabled" with Error Count 0 (point 1) and if TX is started, RX need not be started again.

  • If TX or RX is started consecutively without stopping the earlier run, then difference in the behavior of JNP-MIC1-MACSEC and JNP-MIC1 MICs are observed.

  • Decision feedback equalization (DFE) tuning is required on JNP-MIC1 MIC to start a PRBS test. But, on JNP-MIC1-MACSEC MIC, the DFE tuning is not required. If PRBS is started again at TX or RX without stopping the earlier run, there will be errors until DFE tuning is completed again, in JNP-MIC1 MIC. JNP-MIC1-MACSEC MIC does not show this behavior as there is no DFE tuning involved.

  • You must disable Forward Error Correction (FEC), if loopback is configured in the router with JNP-MIC1 MIC at the remote end.

Clearing the Interface Statistics

The clear interface statistics command clears only the error counters and not the status, RX needs to be restarted to get the right status.

Following are the steps to clear the interface statistics:

  1. Check the statistics at RX by issuing the following command:

    Because only RX is started, there is no PRBS pattern and RX shows maximum error.

  2. To clear the interface statistics, execute the following command:

    Here the status still shows as fail, although the statistics displays the delta value. In this case, because both the current and previous vales are INT_MAX, the delta value 0 is displayed.

    Consider a scenario where the error count increments as the PRBS test is ongoing. In this case, the show interfaces interface-name prbs-stats shows the incremental error (delta value). Also after issuing clear interfaces statistics et-1/1/4 , if the error count is updated, the clear interfaces statistics displays the incremental error too.

    For Example:

    As shown above, in the second instance the show interfaces interface-name prbs-stats command displays the "change" in error count. So, the total error count is 640 + 52 = 692 for Lane 0.