> Subject:
> Re: [44net] Distributed BGP Announce
> From:
> Brian D Heaton <ky9k-lists(a)ky9k.org>
> Date:
> 07/29/2013 05:54 AM
>
> To:
> AMPRNet working group <44net(a)hamradio.ucsd.edu>
>
>
> My experience is with Cisco's "ip tcp adjust-mss xxx". It works 100% in all cases I've seen. Some very large deployed networks use that functionality. The only times we have issues is when folks remove that statement thinking it is redundant to
> MTU size. (We set both)
>
> 73-KY9K/Brian
>
>
> On 7/25/2013 8:59 AM, Marc, LX1DUC wrote
>>> Also as it only really efficts TCP, I solve it on my GRE tunnels with
>>> ip tcp adjust-mss 1436 in cisco
>>> set interface $interface ip tcp adjust-mss 1436 in juniper
>>> tcp-mss-adjust 1436 under an SDP config in Alcatel-Lucent
>>
>> What is your experience with that setup? Does it always (99.999% :-D) work? If so, count me in an let's go with it.
>>
>> 73 de Marc, LX1DUC
I probably invented that :-)
I introduced it into my version of NET (derived from KA9Q NET) in august of 1995, to avoid the
fragmentation that frequently occurred when forwarding IP datagrams over NET/ROM transport.
In my version, the MSS was automatically calculated from the MTU of the incoming and outgoing
interface in the IP routing code. It worked very well. (MTU discovery did not exist back then)
In 2001, Cisco introduced it in IOS and I realized I should have patented it :-)
From changelog:
PE1CHL.950819:
TCP SYN packets are examined when routed, and the MSS option will be
adjusted down to the maximum MSS possible on the incoming and outgoing
interfaces. Thus, a more optimal end-to-end MSS is chosen, and
fragmentation is avoided (e.g. when running IP over NET/ROM somewhere
inbetween the endpoints)
Rob