Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Network Management and Monitoring

  • Timestamp option for tap-aggregation packets (QFX5220-32CD and QFX5220-128C)—High-performance data analysis depends on the accuracy of its underlying data; one accuracy enhancement is a timestamp inserted into a tap-aggregation packet that shows exactly when the packet was captured.

    Configure the tap-aggregation feature to insert a timestamp in packets at data capture, before the packets are sent to the tool ports for analysis. Before the timestamp is inserted, configure the PTP reference clock on the tap-aggregation switch and confirm that PTP is running.. Your tap-aggregation switch must also synchronize the PTP FPGA’s recovered time-of-day with the system chip’s time-of-day. Enable timestamping per interface with

    [edit] user@switch# set interfaces interface-name timestamp ingress

  • Ingress ACL UDF filtering function on tap ports on TAP-aggregation switches (QFX5130-32CD, QFX5130-48C, QFX5220, QFX5230-64CD, QFX5240-64OD, QFX5240-64QD, and QFX5700)—Use an ingress access control list (ACL) with user-defined field (UDFs) on tap interfaces to direct traffic to tool interfaces on a tap-aggregation switch.

    If an ACL match conflicts with a tap-aggregation rule, the ACL match takes precedence.

    Configure the tap-aggregation interfaces under the [edit interfaces] hierarchy level as follows:

    • Add an interface to a tap group with

      [edit]

    • Add an interface to a tool group with

      [edit]

  • Dropped-packet notification (QFX5240-64OD and QFX5240-64QD)—Packet drops commonly occur on network devices, and debugging can be complex and can result in high mean times to recovery (MTTRs). Although the packet-processing pipeline provides some drop counters, they are insufficient for debugging complex packet-drop issues. Use dropped-packet notification (also called mirror on drop, or MoD) to debug drops in real time. This feature monitors packet drops caused due to processing in the ingress and egress pipelines and due to congestion in the memory management unit (MMU). This feature is stateless and flow unaware. Configure most of the feature at the [edit forwarding-options mirror-profile] hierarchy level.