Return-Path: <owner-ips@ece.cmu.edu>
X-Sieve: cmu-sieve 2.0
Return-Path: <owner-ips@ece.cmu.edu>
Received: (from majordom@localhost)
	by ece.cmu.edu (8.11.0/8.10.2) id h0OGlcT20355
	for ips-outgoing; Fri, 24 Jan 2003 11:47:38 -0500 (EST)
X-Authentication-Warning: ece.cmu.edu: majordom set sender to owner-ips@ece.cmu.edu using -f
Received: from d12lmsgate.de.ibm.com (d12lmsgate.de.ibm.com [194.196.100.234])
	by ece.cmu.edu (8.11.0/8.10.2) with ESMTP id h0OGlXW20343;
	Fri, 24 Jan 2003 11:47:33 -0500 (EST)
Received: from d12relay02.de.ibm.com (d12relay02.de.ibm.com [9.165.215.23])
	by d12lmsgate.de.ibm.com (8.12.3/8.12.3) with ESMTP id h0OGlNXP070822;
	Fri, 24 Jan 2003 17:47:23 +0100
Received: from d10ml001.telaviv.ibm.com (d10ml001.telaviv.ibm.com [9.148.216.55])
	by d12relay02.de.ibm.com (8.12.3/NCO/VER6.4) with ESMTP id h0OGlOIo050584;
	Fri, 24 Jan 2003 17:47:24 +0100
In-Reply-To: <15921.25822.312000.475704@gargle.gargle.HOWL>
To: Paul Koning <pkoning@equallogic.com>
Cc: ips@ece.cmu.edu, owner-ips@ece.cmu.edu, pverma@marantinetworks.com
Subject: Re: Min PDU that can come in a TCP packet.
MIME-Version: 1.0
X-Mailer: Lotus Notes Release 6.0 September 26, 2002
From: "Julian Satran" <Julian_Satran@il.ibm.com>
Message-ID: <OF7DFF8105.6F0B2FCE-ONC2256CB8.005B03F2-C2256CB8.005C3A1E@telaviv.ibm.com>
Date: Fri, 24 Jan 2003 18:47:21 +0200
X-MIMETrack: Serialize by Router on D10ML001/10/M/IBM(Release 5.0.9a |January 7, 2002) at
 24/01/2003 18:47:23,
	Serialize complete at 24/01/2003 18:47:23
Content-Type: multipart/alternative; boundary="=_alternative 005B3CC9C2256CB8_="
Sender: owner-ips@ece.cmu.edu
Precedence: bulk

This is a multipart message in MIME format.
--=_alternative 005B3CC9C2256CB8_=
Content-Type: text/plain; charset="US-ASCII"

Paul,

It looks to me like the message:

 >> Is there a possibility that just one byte of an ISCSI PDU can
 >> come in an ISCSI packet?

asks about iSCSI PDUs and in theory even those can have a 1 byte useful 
payload but must be padded

Julo



Paul Koning <pkoning@equallogic.com> 
24/01/03 18:07

To
Julian Satran/Haifa/IBM@IBMIL
cc
pverma@marantinetworks.com, ips@ece.cmu.edu, owner-ips@ece.cmu.edu
Subject
Re: Min PDU that can come in a TCP packet.






>>>>> "Julian" == Julian Satran <Julian_Satran@il.ibm.com> writes:

 Julian> yes but it will have to be padded to 4.

True, but the question wasn't "how small can an iSCSI PDU be" but
rather "can a (possibly large) iSCSI PDU come across in one byte TCP
segments". 

The answer is clearly yes -- TCP is a stream protocol, not a packet
protocol (unfortunately).  There is no required relationship between
TCP segment boundaries and application layer PDU boundaries.  An
application layer PDU may come across as one TCP segment; you may have
multiple application PDUs in one TCP segment; application PDUs may
cross TCP segment boundaries; or an application PDU may come across in
multiple TCP segments, theoretically in units as small as one byte per
TCP segment.

In practice I would assume that sensible implementers will not, or not
often, do some of the more extreme things permitted by TCP, but if you
want a reliabile implementation, you have to permit all sorts of
strange things.

                 paul

 >> Is there a possibility that just one byte of an ISCSI PDU can
 >> come in an ISCSI packet?

 >> Preetam Verma



--=_alternative 005B3CC9C2256CB8_=
Content-Type: text/html; charset="US-ASCII"


<br><font size=2 face="sans-serif">Paul,</font>
<br>
<br><font size=2 face="sans-serif">It looks to me like the message:</font>
<br>
<br><font size=2><tt>&nbsp;&gt;&gt; Is there a possibility that just one
byte of an ISCSI PDU can<br>
 &gt;&gt; come in an </tt></font><font size=2 color=red><tt>ISCSI packet</tt></font><font size=2><tt>?</tt></font>
<br><font size=2><tt><br>
</tt></font><font size=2 face="sans-serif">asks about iSCSI PDUs and in
theory even those can have a 1 byte useful payload but must be padded</font>
<br>
<br><font size=2 face="sans-serif">Julo</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>Paul Koning &lt;pkoning@equallogic.com&gt;</b>
</font>
<p><font size=1 face="sans-serif">24/01/03 18:07</font>
<td width=59%>
<table width=100%>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">To</font></div>
<td valign=top><font size=1 face="sans-serif">Julian Satran/Haifa/IBM@IBMIL</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td valign=top><font size=1 face="sans-serif">pverma@marantinetworks.com,
ips@ece.cmu.edu, owner-ips@ece.cmu.edu</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Subject</font></div>
<td valign=top><font size=1 face="sans-serif">Re: Min PDU that can come
in a TCP packet.</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=2><tt>&gt;&gt;&gt;&gt;&gt; &quot;Julian&quot; == Julian
Satran &lt;Julian_Satran@il.ibm.com&gt; writes:<br>
<br>
 Julian&gt; yes but it will have to be padded to 4.<br>
<br>
True, but the question wasn't &quot;how small can an iSCSI PDU be&quot;
but<br>
rather &quot;can a (possibly large) iSCSI PDU come across in one byte TCP<br>
segments&quot;. &nbsp;<br>
<br>
The answer is clearly yes -- TCP is a stream protocol, not a packet<br>
protocol (unfortunately). &nbsp;There is no required relationship between<br>
TCP segment boundaries and application layer PDU boundaries. &nbsp;An<br>
application layer PDU may come across as one TCP segment; you may have<br>
multiple application PDUs in one TCP segment; application PDUs may<br>
cross TCP segment boundaries; or an application PDU may come across in<br>
multiple TCP segments, theoretically in units as small as one byte per<br>
TCP segment.<br>
<br>
In practice I would assume that sensible implementers will not, or not<br>
often, do some of the more extreme things permitted by TCP, but if you<br>
want a reliabile implementation, you have to permit all sorts of<br>
strange things.<br>
<br>
 &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;
paul<br>
<br>
 &gt;&gt; Is there a possibility that just one byte of an ISCSI PDU can<br>
 &gt;&gt; come in an ISCSI packet?<br>
<br>
 &gt;&gt; Preetam Verma<br>
<br>
</tt></font>
<br>
--=_alternative 005B3CC9C2256CB8_=--
