Fix on howto increase the number of IPv6 (BGP) route entries in HW LPM

EtherNation_Use — Wed, 08 Jan 2014 05:59:00 GMT

Create Date: Mar 10 2013 1:44PM

Hi,

As some of you may have seen, the current IPv6 BGP Tablesize is growing quickly currently.

That might give messages in your LOG on the switch stating that it can't download a route into HW LPM. ( HW LPM table full )

In the latest 15.2.2.7 releases (and up), there is a fix to make a change in the distribution of the current hw entries between IPv4 and IPv6 L3 entries.

The command : configure forwarding external-tables l3-only ipv4-and-ipv6 (*reboot required to activate)

This will provide more room for IPv6 prefixes (upto 47.5k routes) by lowering the IPv4 HW LPM to 475k entries.

To state the obvious, you should always have both IPv4 and IPv6 route compression enabled.

enable iproute ipv4 compression

enable iproute ipv6 compression

This will result in a similar route stat (If you have both IPv4 and IPv6 running on BGP on an X480)

sho iproute reserved-entries statistics

|-----In HW Route Table----| |--In HW L3 Hash Table--|

# Used Routes # IPv4 Hosts IPv4 IPv4 IPv6 IPv4

Slot Type IPv4 IPv6 Local Remote Local Rem. Loc. MCast

---- ---------------- ------ ------ ------ ------ ----- ----- ---- -----

1 X480-24x 242489 5794> 41 0 0 0 4 22

sho iproute summary

=================ROUTE SUMMARY=================

Mask distribution:

2 default routes 17 routes at length 8

13 routes at length 9 29 routes at length 10

88 routes at length 11 247 routes at length 12

481 routes at length 13 863 routes at length 14

1554 routes at length 15 12517 routes at length 16

6568 routes at length 17 10920 routes at length 18

21563 routes at length 19 31024 routes at length 20

33172 routes at length 21 45101 routes at length 22

40705 routes at length 23 231077 routes at length 24

81 routes at length 25 75 routes at length 26

74 routes at length 27 85 routes at length 28

75 routes at length 29 158 routes at length 30

159 routes at length 32

Route origin distribution:

436181 IBGP 450 OSPFExt1 1 OSPFInter

8 OSPFIntra 2 Blackhole 3 Static

3 Direct

Total number of routes = 436648

Total number of compressed routes = 194052

For IPv6:

sho bgp rou address-family ipv6-unicast summary

Flags: (*) Preferred BGP route, (>) Active, (d) Suppressed, (h) History

(s) Stale, (m) Multipath, (u) Unfeasible

Origin: (?) Incomplete, (e) EGP, (i) IGP

BGP Route Statistics

Total Rxed Routes : 11667

Feasible Routes : 11667

Active Routes : 11667

Rejected Routes : 0

Unfeasible Routes : 0

Advertised Routes : 5

Route Statistics on Session Type

Routes from Int Peer: 0

Routes from Ext Peer: 11667

For IPv4 :
sho bgp rou sum

Flags: (*) Preferred BGP route, (>) Active, (d) Suppressed, (h) History

(s) Stale, (m) Multipath, (u) Unfeasible

Origin: (?) Incomplete, (e) EGP, (i) IGP

BGP Route Statistics

Total Rxed Routes : 489325

Feasible Routes : 489325

Active Routes : 436125

Rejected Routes : 0

Unfeasible Routes : 0

Advertised Routes : 48

Route Statistics on Session Type

Routes from Int Peer: 489325

Routes from Ext Peer: 0

GL,
Erik Bais
A2B Internet

(from Erik_Bais)

RE: Fix on howto increase the number of IPv6 (BGP) route entries in HW LPM

EtherNation_Use — Wed, 08 Jan 2014 05:59:00 GMT

Create Date: Mar 11 2013 12:07AM

Pretty cool - I hadn't seen that in the release notes, great find! Thanks! (from Ansley_Barnes)

topic Fix on howto increase the number of IPv6 (BGP) route entries in HW LPM in ExtremeSwitching (EXOS/Switch Engine)

Fix on howto increase the number of IPv6 (BGP) route entries in HW LPM

RE: Fix on howto increase the number of IPv6 (BGP) route entries in HW LPM