> On Jul 29, 2014, at 12:03 AM, "Zimmermann, Alexander" > <[email protected]> wrote: > > Hi Joe, > >> Am 28.07.2014 um 17:37 schrieb Joe Touch <[email protected]>: >> >> >> >>> On 7/28/2014 1:38 AM, Zimmermann, Alexander wrote: >>> Hi, >>> >>>> Am 28.07.2014 um 08:57 schrieb marcelo bagnulo braun <[email protected]>: >>>> >>>> Hi, >>>> >>>> As we discussed in the meeting, we should try to make some design >>>> decisions for TCPINC. >>>> One of them is whether to protect or not the TCP header. >>>> I would like to start the discussion on this topic. Arguments on one way >>>> or the other? >>> >>> if folks want to protect the TCP header, I don’t see a (straightforward) >>> way on how we can >>> use todays TOE. No TOE is a NO-GO for (at least) data center environments. >> >> TOE doesn't replace the OS TCP, or the fact that the OS TCP is in control of >> the offload support. For any TCP modification, either the TOE supports that >> modification with an extended interface to the OS or not; if not, then the >> OS can't use the TOE for that connection. >> >> None of this is new or specific to TCPINC. > > full ACK. However, how many TCP extension we develop in the last decade that > breaks TOE? > See for example MPTCP. You can design an extension in that way that doesn’t > break TOE.
The same is true for some of the proposed solutions. AO-enc e.g. Joe > >> >> Joe >> >> _______________________________________________ >> Tcpinc mailing list >> [email protected] >> https://www.ietf.org/mailman/listinfo/tcpinc > _______________________________________________ Tcpinc mailing list [email protected] https://www.ietf.org/mailman/listinfo/tcpinc
