Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Community Member

full BGP table balancing

Hi

we have a client connected to two service provider routers through two links ( every link to every SP router). Every SP router have the full BGP internet table. We'd like to send the full BGP table to the client but we'd like to balance the traffic through both links. Could you suggest me the best way to do that? Tks

2 ACCEPTED SOLUTIONS

Accepted Solutions
Hall of Fame Super Silver

Re: full BGP table balancing

Hello Pedro,

taking in account the scalability issues.

a)

router bgp xx

maximum-paths 2

easy to configure but it can be too heavy at level of resource usage

b)

another approach I've seen in the forums is

create two prefix-lists with each that matches one half of IPv4 address space :

provide a bigger local-preference to ISP1 for first half of ipv4 address space and a bigger local preference to ISP2 for the other half of IPv4 address space

prefix-list second-half permit ip 128.0.0.0/1 le 24

route-map ISP1 permit 10

match ip address prefix-list second-half

set local-preference 200

route-map ISP1 permit 20

set local-preference 120

route-map ISP2 permit 10

set local-preference 150

router bgp

neigh isp1-ipaddr route-map ISP1 in

neigh isp2-ipaddr route-map ISP2 in

Don't expect a fair load sharing with this second method and also return path ratio can be totally different.

Hope to help

Giuseppe

Cisco Employee

Re: full BGP table balancing

Giuseppe,

The first option will not work by default if the two SPs are in different ASes. This is because maximum-path only works for prefixes received from the same neighbor AS by default.

You would need the following hidden command to change this default behavior:

bgp bestpath as-path multipath-relax

This command (along with the maximum-paths x) would cause BGP to select both paths even if they come from different neighbor AS, assuming all other attributes are equals.

Regards

Harold Ritter
Sr. Technical Leader
CCIE 4168 (R&S, SP)
harold@cisco.com
México móvil: +52 1 55 8312 4915
Cisco México
Paseo de la Reforma 222
Piso 19
Cuauhtémoc, Juárez
Ciudad de México, 06600
México
3 REPLIES
Hall of Fame Super Silver

Re: full BGP table balancing

Hello Pedro,

taking in account the scalability issues.

a)

router bgp xx

maximum-paths 2

easy to configure but it can be too heavy at level of resource usage

b)

another approach I've seen in the forums is

create two prefix-lists with each that matches one half of IPv4 address space :

provide a bigger local-preference to ISP1 for first half of ipv4 address space and a bigger local preference to ISP2 for the other half of IPv4 address space

prefix-list second-half permit ip 128.0.0.0/1 le 24

route-map ISP1 permit 10

match ip address prefix-list second-half

set local-preference 200

route-map ISP1 permit 20

set local-preference 120

route-map ISP2 permit 10

set local-preference 150

router bgp

neigh isp1-ipaddr route-map ISP1 in

neigh isp2-ipaddr route-map ISP2 in

Don't expect a fair load sharing with this second method and also return path ratio can be totally different.

Hope to help

Giuseppe

Cisco Employee

Re: full BGP table balancing

Giuseppe,

The first option will not work by default if the two SPs are in different ASes. This is because maximum-path only works for prefixes received from the same neighbor AS by default.

You would need the following hidden command to change this default behavior:

bgp bestpath as-path multipath-relax

This command (along with the maximum-paths x) would cause BGP to select both paths even if they come from different neighbor AS, assuming all other attributes are equals.

Regards

Harold Ritter
Sr. Technical Leader
CCIE 4168 (R&S, SP)
harold@cisco.com
México móvil: +52 1 55 8312 4915
Cisco México
Paseo de la Reforma 222
Piso 19
Cuauhtémoc, Juárez
Ciudad de México, 06600
México
Hall of Fame Super Silver

Re: full BGP table balancing

Hello Harold,

you are right and you had already made the same correction some mounths ago together with Narayan to another post of mine.

Thanks

Best Regards

Giuseppe

125
Views
0
Helpful
3
Replies
CreatePlease to create content