Re: Urgent as Framing Hint?

["Randall R. Stewart" <randall@stewart.chicago.il.us>]

John:

I think you got a couple of things mixed up..

The person you quote below is NOT me.. it is David Reed.
Now the "magic number" method is to imbed a framing sequence
within the data stream, with or without a escape sequence.

What David was proposing is twiddling the TCP checksum so
that iSCSI start frames have a "different" special c-sum.
It is a quick and kinda elegant hack, if a hack can
be considered elegant... and I did like it... however has
you point out there are drawbacks, i.e. a NAT would barf on
this :)

Lets term David's proposal the "checksum twiddle" so we can
keep this seperate from the "magic number" proposed in one of
the drafts and previously discussed on the ips list (only .. not
the end2end list).

R


John H Dowdy wrote:
> 
> Randall liked the idea of the magic number. The following statement
> concerns me, though:
> > I personally like this idea a lot, because any "middleboxes" that screw
> > with checksums (even to check them) or translate streams will be
> discovered
> > and shamed publicly for their egregious violation of layering.
> 
> Won't this be yet another way for NAT boxes to break iSCSI sessions? I.e.,
> a NAT box will change
> addresses AND TCP checksums - typically, they delta the checksum based on
> the changes made
> to IP addresses. Nevertheless, won't any change to the magic number break
> it's usefullness?
> 
> John Dowdy - SSG Development
> Telephone (919)254-5632 (T/L 444-5632)
> jdowdy@us.ibm.com

-- 
Randall R. Stewart
randall@stewart.chicago.il.us or rrs@cisco.com
815-342-5222 (cell) 815-477-2127 (work)