Multilink PPP differences

Have a question or want to start a discussion? Post it! No Registration Necessary.  Now with pictures!

Threaded View
Hello,

I am wonder what the difference between the two configurations below
are. It seems there are two ways to configure MLPPP. Either using a
multilink interface or a virtual template. One uses an unnumbered IP
and one uses a /30 for the multilink interface. I see in the QOS config
guids that you use option #2 for MLPPP interleave.

Which option is best practice for normal circumstances?

****Configuration #1****
multilink virtual-template 1
!
interface Virtual-Template1
 ip unnumbered Loopback0
 ppp authentication chap
 ppp multilink
!
interface Serial0/0/0:1
 no ip address
 encapsulation ppp
 no fair-queue
 ppp multilink
!
interface Serial0/0/1:1
 no ip address
 encapsulation ppp
 no fair-queue
 ppp multilink


****Configuration #2****
interface Multilink1
 ip address 10.xxx.xxx.xxx 255.xxx.xxx.xxx
 ppp multilink
 ppp multilink group 1
!
interface Serial0/0/0
 no ip address
 ppp multilink
 ppp multilink group 1
 encapsulation ppp
 no fair-queue
!
interface Serial0/1/0
 no ip address
 ppp multilink
 ppp multilink group 1
 encapsulation ppp
 no fair-queue


Re: Multilink PPP differences
Quoted text here. Click to load it

Since Multilink is statically configured, it's well suitable for leased-line
style connections. Virtual-Template is normally used in dial-like
environments on provider side where you have many users that need to share
similar configuration and you don't want to (or simply can't) configure each
user link individually. You can use PPP LFI with either of this approaches.
Also none of this scenarios forces you to only one addressing strategy - you
can use /30 with Virtual-Template (we do for some users) and you can use
unnumbered for Multilink interface.

Kind regards,
iLya



Site Timeline