From: John Gibson (johngibson1541@yahoo.com)
Date: Thu May 24 2007 - 15:10:42 ART
Hi Kenny, do SPs typically prevent mtu discovery
from outside ? (some filtering)
I just keep getting the impression that SPs don't
like ICMP from outside probing the internal.
John
--- spduo <frenzeus@streamyx.com> wrote:
> Hi John,
>
> it is still very much used in bgp environment AFAIK,
> especially in SP
> environment. As BGP uses TCP, and by default every
> TCP session has a limit
> in terms of how much data it can transport in a
> single packet, whilst this
> limit also known as the Maximum Segment Size (MSS)
> on the IOS routers is 536
> bytes only. When using such a small value, it
> creates a large amount of
> tcp/ip overhead, and especially when tcp carries a
> lot of data when bgp is
> using it! tcp will take all the big chunks of data
> and break it up into 536
> bytes chunks, by default.
>
> With tcp path mtu discovery, it allows tcp to
> determine the smallest mtu
> size among all links between the ends of a tcp
> session, i.e. ur ibgp
> neighbor peering btw PE-PE in MPLSVPN. Usually there
> are multiple hops (P
> routers) in between the 2 PE routers, and hence the
> need to enable "ip tcp
> path-mtu-discovery".
>
> do a "show ip bgp nei | i max data" before and after
> you configure "ip tcp
> path-mtu-discovery" and u'll see wut i mean. the
> values displayed, after u
> configure this.. is actually ur smallest mtu along
> the path of ur ibgp
> session, minus of ip and tcp headers.
>
> HTH.
>
> Kenny
>
> ----- Original Message -----
> From: <johngibson1541@yahoo.com>
> To: <ccielab@groupstudy.com>
> Sent: Thursday, May 24, 2007 6:04 AM
> Subject: path mtu discovery is considered legacy?
>
>
> >
>
This archive was generated by hypermail 2.1.4 : Fri Jun 01 2007 - 06:55:22 ART