Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
ON THIS PAGE
 

Detect Physical Layer Faults

Use this topic to understand how Juniper Routing Director automatically detects physical layer faults (bad or faulty cables) and how you can use the GUI to view alerts and information related to the faults.

Physical Layer Fault Detection Overview

When a network cable is unplugged or breaks, it is easy for an operator to pinpoint the issue because all traffic transmitted through the cable drops. However, when a cable degrades or is not plugged in properly, the signal carried by the cable is attenuated or weakened. This can cause some traffic through the cable to drop. Because the cable is partially functional, it is difficult for operators to detect that the traffic drops are caused by a faulty cable.

Routing Director enables you to automatically detect faults in cables (also called bad cables or physical layer faults) by using AI/ML (artificial intelligence [AI] and machine learning [ML]) techniques. Routing Director uses trained ML models to analyze the data received from network devices and detects when a cable has turned faulty. It then raises an alert in the GUI, which enables you to easily identify the faulty cables. You can replace such faulty cables before they cause traffic disruption.

Routing Director automatically detects physical layer faults on 100G CWDM and LR4 optical cables connected to ACX7100-48L.

Routing Director detects physical layer faults in the following scenarios:

  • During device onboarding—When a device is being onboarded, Routing Director can detect if a cable connected to the device is not working properly, and then alert the field technician advising a replacement. However, because this detection takes place during device onboarding, when the device has a short history, the accuracy of the prediction is limited. During device onboarding, Routing Director triggers an alert for faulty cables in approximately 10 minutes after the neighbor ping test is initiated.

  • During device operation—After the device is onboarded successfully and is managed, Routing Director monitors the device continuously. It then uses the historical data to detect physical layer faults, and triggers an alert within 30 minutes of the cable turning faulty.

Physical Layer Fault Notifications in the GUI

In the Routing Director GUI, physical layer fault notifications for a device are shown on the Connectivity accordion of the Device-Name page as shown in Figure 1.

Figure 1: Physical Layer Fault Alert in Connectivity Accordion Physical Layer Fault Alert in Connectivity Accordion

To view physical layer fault notifications in the GUI:

  1. Do one of the following:

    • To view physical layer fault notifications during device onboarding, select Inventory > Device Onboarding > Onboarding Dashboard > Put Devices into Service > Device-Name.

    • To view physical layer fault notifications during device operation, select Observability > Troubleshoot Devices > Device-Name.

    The Device-Name page appears.

  2. Scroll to the Connectivity accordion and click > to expand the accordion.

    Physical layer fault event notifications appear under Relevant Events with the following information:

    • Event notification message in the format Physical Layer Fault of type cable-type cable-info detected in interface-name, where:

      • cable-type is the type of cable: optical.

      • cable-info is additional info about the cable; for example, LR4.

      • interface-name is the name of the device interface to which the cable is connected. The interface name is in the format interface-short-form-fpc-slot/pic-slot/port-number; for example, et-0/0/1.

    • Date and time the event was triggered. Hover over the date and time to view the last time an alert for the fault was generated (Last Received Time).

    An example of a physical layer fault event notification is Physical Layer Fault of type Optical LR4 detected in interface et-0/0/48.

    The physical layer fault is also displayed as an alert on the Events page (Observability > Health > Events) as a Connection Switch type event.

    If the alert is not acknowledged and additional physical layer faults are not detected for a 24-hour period, the notification is cleared automatically.

  3. Hover over or click View Details to view the details of the event, including the end date and time for the fault event.
  4. (Optional) Click View All Relevant Events to view all the connectivity-related events for the device, including physical layer fault events.

    The events are listed as alerts on the Events page (Observability > Health > Events.) as shown in figure.

    Figure 2: Alerts Indicating Physical Layer Faults Alerts Indicating Physical Layer Faults

See Also