Configuring IPv4 and IPv6 Accounting

IPv4 and IPv6 accounting are disabled by default. But you can enable the accounting by including the route-accounting statements at the [edit forwarding-options family inet4] and [edit forwarding-options family inet6] hierarchy level, as shown here:

To view the IPv4 or IPv6 statistics for a given physical or logical interface, run the operational command show interfaces extensive interface | find IPv4 or show interfaces extensive interface | find IPv6. Note that the output displays packet and byte counts for transit traffic only. Locally generated packets are not included in the metrics.

Follow the guidelines given below to configure IPv4 and IPv6 statistics accounting:

• The transit rate is calculated in software and not in hardware. So, you can expect some deviation from the line rate. You do not receive the expected rate with lower fps (for example, 5 fps).

• The traffic must be IPv4 or IPv6 at both ingress and egress. Any form of encapsulated traffic or traffic translating to IPv4 is not supported on PTX routers.

• If the GRE tunnel is IPv4 or IPv6 based, this traffic is not accounted in ingress statistics.

• Statistics accounting is supported only for et and ae interfaces on PTX routers.

• The virtual, software, pseudo, Layer 2, and special type Layer 3 virtual interfaces are not supported. For example, interfaces such as the following:

  • irb

  • rvi

  • lsi

  • fti

  • vtep

  • vt

• IPv4 transit statistics is not displayed on Routine Engine show commands.

• No interoperability between SCU and IPv4 statistics accounting on PTX routers.

  The following functionalities are not supported:

  • lo0 traffic

  • IPv4 statistics query over SNMP

  • Aggregated Ethernet child link IPv4 statistics accounting