Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.protocols.tcp-ip    |    TCP and IP network protocols.    |    14,669 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 13,227 of 14,669    |
|    Arlet to Steinar Haug    |
|    Re: When are payload lengths greater tha    |
|    28 Nov 09 15:25:27    |
      From: usenet+6@ladybug.xs4all.nl              On Sat, 28 Nov 2009 13:45:27 +0100, Steinar Haug wrote:              >> >>> Data payloads with length greater than the MSS negotiated are being       >> >>> exchanged. I would like to know under which conditions are payloads       >> >>> greater than the MSS exchanged? Further, is there a way to limit the       >> >>> payloads to the specified MSS value?       >> >>       >> >> I've seen this happening with NICs that do tcp offloading. In that case,       >> >> the frame you see in tcpdump is NOT what goes over the wire, but what the       >> >> OS send to the card. FWIW, this was on AIX.       >> >       >> > The original post states that this occurred on the loopback device       >>       >> Linux has TSO enabled on the loopback device since version 2.6.27.       >>       >> # ethtool -k lo       >> [...]       >> tcp segmentation offload: on       >       > Um, what exactly do you offload this processing to? Since the loopback       > interface is internal on the machine (internal to the operating system       > if you wish), I don't really see the point of TSO ...              It would prevent TCP segmentation by just copying the application data       from sender to receiver.              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca