Re: [tcpm] TCP persist state issue.

On Mon, Jun 30, 2008 at 10:44 PM, Anantha Ramaiah (ananth)
<ananth@cisco.com> wrote:
>> You can find the copy of the first draft at :
>> http://www.ietf.org/internet-drafts/draft-ananth-tcpm-persist-00.txt.
>> Comments welcome.

As with the previous draft, I think the greatest weakness of this one
is that it focuses so much on the persist state.  The conclusion
states, "The document addresses the fact that terminating TCP
connections stuck in the persist condition does not violate any RFC."
But isn't this statement also true without the condition that
connections are "stuck" in persist?  That is, terminating a connection
at any time is allowed, regardless of its state (see RFC 793, section
3.8, the ABORT command).

And again, I think the persist state has fairly little to do with the
more general DoS issue.  Consider the following: a client that,
instead of shrinking its window to zero, simply stops ACKing after
sending the request, or a client that slowly acks one byte at a time.
Both have the same effect on the server, are fairly straightforward,
and neither advertise a zero window.

  -John
_______________________________________________
tcpm mailing list
tcpm@ietf.org
https://www.ietf.org/mailman/listinfo/tcpm