[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: FullTcp and DelAck


> [*] Mark Allman turns that into a 'must' on the first page;
>     did I miss an RFC rev?
> http://www.acm.org/sigcomm/ccr/archive/1998/oct98/ccr-9810-allman.html

RFC 1122 is ambiguous on the use of delayed ACKs.  The paper cited
above does say that delayed ACKs are optional and if used a receiver
must ACK every second incoming segment.  I *believe* this is a true
statement if you read one section of 1122 and false if you read
another section.

We cleaned this up a bit in RFC 2581 (which appears after the above
cited CCR paper), as follows:

  * delayed ACKs SHOULD be used by receivers
  * when using delayed ACKs, a receiver MUST NOT excessivly delay
    ACKs, i.e.,
      * an ACK SHOULD be sent for every second full sized segment
	that arrives
      * an ACK MUST be sent within 500~ms of the first data segment
        arrival 

allman


---
http://roland.grc.nasa.gov/~mallman/