Increased IP Route Scaling with Algorithmic Longest-Prefix Match (ALPM)

This feature increases IP route scaling in hardware using algorithmic longest-prefix match (ALPM). You can now configure flexible internal hardware lookup tables to hold more IP routes, in addition to previous options to hold more MAC addresses or more IP unicast and multicast entries. With ALPM configured, additional IPv4 routes and IPv6 routes with subnet mask length less than or equal to 64 bits are stored in hardware for forwarding IP unicast packets (see Limits). The algorithmic capacity is not fixed, and depends on the mix of IP subnets and mask lengths in use:
  • IPv4 route scale increases from 16K:
    • For Summit X670-G2 and X770: minimum of 128K and maximum of 256K.
    • For ExtremeSwitching X870: minimum of 128K and maximum of 192K.
  • IPv6 route scale increases from 8K for subnet mask length 0–64:
    • Summit X670-G2 and X770: minimum of 84K and a maximum of 128K.
    • For ExtremeSwitching X870: 64K
  • ECMP for IPv4 and IPv6 routes, user VRs and inter-VR routing for IPv4 are still supported.

Supported Platforms

Summit X670-G2, X770, and ExtremeSwitching X870 series switches.

Changed CLI Commands

Changes are underlined.

configure forwarding internal-tables [ l2-and-l3 | more [l2 | l3-and-ipmc | routes] ]