Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 23 May 1998 20:49:06 +0000 (GMT)
From:      Terry Lambert <tlambert@primenet.com>
To:        njs3@doc.ic.ac.uk (Niall Smart)
Cc:        hackers@FreeBSD.ORG
Subject:   Re: tcp states and sysctl's
Message-ID:  <199805232049.NAA12898@usr07.primenet.com>
In-Reply-To: <E0ycyPn-0002lv-00@oak66.doc.ic.ac.uk> from "Niall Smart" at May 22, 98 09:27:42 pm

next in thread | previous in thread | raw e-mail | index | archive | help
> I don't think a shutdown(2) sysctl is necessary.
> 
> a) the BSD stack is currently compliant in this regard, if
>    anything need's changing it's the client.
> 
> b) no-one has produced any evidence to show that all
>    these sockets in TIME_WAIT_2 are actully having a
>    negative impact in performance on the system.  Actually,
>    I would seriously hope not, because otherwise this
>    is a relatively easy DoS.  (Though probably not as
>    effective as a SYN flood.)
>  
> If you do decide to put in a sysctl bear in mind that rsh uses
> shutdown(2) to close down one end of a socket so you don't want
> the timeout to be too short.

Actually, there is a bug in zero-window probing.  This occurs when
talking to an HP Laserwriter.

Basically, the FreeBSD box closes the connection, and can't send
a FIN because the printer announces a zero window.

A mechanism for converting a shutdown 1 to a shutdown 2 would be
useful here.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199805232049.NAA12898>