Each computer knows the MTU between it and the next box. The gateway sends the
MTU of the ppp link as part of its request so the remote server uses that size
(unless the server has an even smaller link, which is unlikely). The
workstation only knows the MTU to the gateway (1500). Until the ICMP packets
come in, neither the workstation nor the server realize there's a constricted
link along the way.
Do load balancers and IP masq pass ICMP packets along to the inside box?
Without knowing much about masq, my guess would be no. I assume they both use
address:port->address:port relationships to forward packets to the proper
destination and ICMP doesn't have port information.
On Thu, 18 May 2000, you wrote:
> I see, so what is happening is someone along the way is requesting
> fragmentation, so we receive small packets that fit into our narrow MTU
> tube. Sites block an ICMP 'please frag' and thus send back larger
> packets.
>
> I remain unable to understand why my gateway is able to connect fine, and
> why _it_ doesn't defrag them so we can receive them internally. Changing
> MTU on the ppp link fixed the problem.
>
> Am I understanding this properly? (Should I sit down and read Steven's UNP
> and Internetworking again?) Isn't there anything in /proc/sys/net/ipv4
> that would adjust the defrag behaviour? Or is stuff getting lost before my
> end?
>
> Cheers,
> --
> /\ Christian Reis exists solely to answer [EMAIL PROTECTED]
> \'`/ Async Free SW Development | http://async.com.br | +55 16 274 2497
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-net" in
> the body of a message to [EMAIL PROTECTED]
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
