The documentation says that Path MTU discovery is enabled by default for all 
BGP neighbor sessions:

http://www.cisco.com/en/US/docs/ios/iproute/configuration/guide/irp_bgp_neighor.html#wp1054165

So enabling "ip tcp path-mtu-discovery" should not be necessary.

But i made some tests with 12.0S and it seems the documentation is not accurate:

- In the same 1500 bytes MTU ethernet segment, without that command the SYN 
packet is sent with MSS=536 bytes.
- With that command, the SYN packet is sent with MSS=1460 bytes.

This seems buggy to me :(


Regards,

Antonio Soares, CCIE #18473 (R&S)
[email protected]

-----Original Message-----
From: [email protected] 
[mailto:[email protected]] On Behalf Of Lukasz Bromirski
Sent: quarta-feira, 11 de Fevereiro de 2009 23:40
To: Jose Conceicao
Cc: cisco-nsp
Subject: Re: [c-nsp] BGP MSS=576 bytes

On 2009-02-11 21:51, Jose Conceicao wrote:
> After choosing whatever prefered method, wouldn't these BGP sessions 
> need to be re-established to take effect? (since MSS is only carried 
> in SYN and SYN/ACKs)

Sure.

> And shouldn't u account for TCP/IP overhead? (just being lazy by not 
> testing this in lab)

TCP/IP overhead of what actually?

Here's a whitepaper on PMTUD/MSS:
http://www.cisco.com/en/US/tech/tk827/tk369/technologies_white_paper09186a00800d6979.shtml

And there's CSCef52888 for all of You pre 12.2(18)SXF12 guys which doesn't 
actually have PMTUD do it's work.

-- 
"Don't expect me to cry for all the     |               Łukasz Bromirski
  reasons you had to die" -- Kurt Cobain |    http://lukasz.bromirski.net
_______________________________________________
cisco-nsp mailing list  [email protected] 
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

_______________________________________________
cisco-nsp mailing list  [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Reply via email to