cbiedl
/
tcpreplay


			
				
					
						
						
							123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282
							<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">

<!--Converted with LaTeX2HTML 2002-2 (1.70)
original version by:  Nikos Drakos, CBLU, University of Leeds
* revised and updated by:  Marcus Hennecke, Ross Moore, Herb Swan
* with significant contributions from:
  Jens Lippmann, Marek Rouchal, Martin Wilck and others -->
<HTML>
<HEAD>
<TITLE>4 Multiple Independent Flows</TITLE>
<META NAME="description" CONTENT="4 Multiple Independent Flows">
<META NAME="keywords" CONTENT="flowreplay">
<META NAME="resource-type" CONTENT="document">
<META NAME="distribution" CONTENT="global">

<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="LaTeX2HTML v2002-2">
<META HTTP-EQUIV="Content-Style-Type" CONTENT="text/css">

<LINK REL="STYLESHEET" HREF="flowreplay.css">

<LINK REL="next" HREF="node5.html">
<LINK REL="previous" HREF="node3.html">
<LINK REL="up" HREF="flowreplay.html">
<LINK REL="next" HREF="node5.html">
</HEAD>

<BODY >

<DIV CLASS="navigation"><!--Navigation Panel-->
<A NAME="tex2html72"
  HREF="node5.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next" SRC="next.png"></A> 
<A NAME="tex2html70"
  HREF="flowreplay.html">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up" SRC="up.png"></A> 
<A NAME="tex2html64"
  HREF="node3.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous" SRC="prev.png"></A>   
<BR>
<B> Next:</B> <A NAME="tex2html73"
  HREF="node5.html">5 pcap vs flow</A>
<B> Up:</B> <A NAME="tex2html71"
  HREF="flowreplay.html">Flowreplay Design Notes</A>
<B> Previous:</B> <A NAME="tex2html65"
  HREF="node3.html">3 Design Thoughts</A>
<BR>
<BR></DIV>
<!--End of Navigation Panel-->
<!--Table of Child-Links-->
<A NAME="CHILD_LINKS"><STRONG>Subsections</STRONG></A>

<UL CLASS="ChildLinks">
<LI><A NAME="tex2html74"
  HREF="node4.html#SECTION00041000000000000000"><SPAN CLASS="arabic">4</SPAN>.<SPAN CLASS="arabic">1</SPAN> <SPAN ID="hue158">IP Fragments and TCP Streams</SPAN></A>
<LI><A NAME="tex2html75"
  HREF="node4.html#SECTION00042000000000000000"><SPAN CLASS="arabic">4</SPAN>.<SPAN CLASS="arabic">2</SPAN> <SPAN ID="hue194">Blocking</SPAN></A>
</UL>
<!--End of Table of Child-Links-->
<HR>

<H1><A NAME="SECTION00040000000000000000">
<SPAN CLASS="arabic">4</SPAN> <SPAN ID="hue142">Multiple Independent Flows</SPAN></A>
</H1>

<P>
<SPAN ID="hue144">The biggest asynchronous problem, that pcap files
are serial, has to be solved in a scaleable manner. Not much can be
assumed about the network traffic contained in a pcap savefile other
then Murphy's Law will be in effect. This means we'll have to deal
with:</SPAN>
<P>

<UL>
<LI><SPAN ID="hue147">Thousands of small simultaneous flows (captured
on a busy network)</SPAN>
</LI>
<LI><SPAN ID="hue380">Flows which ``hang'' mid-stream (an exploit
against a server causes it to crash)</SPAN>
</LI>
<LI><SPAN ID="hue151">Flows which contain large quantities of data (FTP
transfers of ISO's for example)</SPAN>
</LI>
</UL>
<SPAN ID="hue154">How we implement parallel processing of the pcap
savefile will dramatically effect how well we can scale. A few considerations:</SPAN>
<P>

<UL>
<LI>Most Unix systems limit the maximum number of open file descriptors
a single process can have. Generally speaking this shouldn't be a
problem except for highly parallel pcap's.
</LI>
<LI>While RAM isn't limitless, we can use mmap() to get around this.
</LI>
<LI>Many Unix systems have enhanced solutions to poll() which will improve
flow management.
</LI>
</UL>

<P>

<H2><A NAME="SECTION00041000000000000000">
<SPAN CLASS="arabic">4</SPAN>.<SPAN CLASS="arabic">1</SPAN> <SPAN ID="hue158">IP Fragments and TCP Streams</SPAN></A>
</H2>

<P>
<SPAN ID="hue160">There are five major complications with flowreplay:</SPAN>
<P>

<OL>
<LI><SPAN ID="hue163">The IP datagrams may be fragmented- we won't be
able to use the standard 5-tuple (src/dst IP, src/dst port, protocol)
to lookup which flow a packet belongs to.</SPAN>
</LI>
<LI><SPAN ID="hue165">IP fragments may arrive out of order which will
complicate ordering of data to be sent.</SPAN>
</LI>
<LI><SPAN ID="hue167">The TCP segments may arrive out of order which will
complicate ordering of data to be sent.</SPAN>
</LI>
<LI><SPAN ID="hue169">Packets may be missing in the pcap file because
they were dropped during capture.</SPAN>
</LI>
<LI><SPAN ID="hue171">There are tools like fragrouter which intentionally
create non-deterministic situations.</SPAN>
</LI>
</OL>
<SPAN ID="hue174">First off, I've decided, that I'm not going to worry
about fragrouter or it's cousins. I'll handle non-deterministic situations
one and only one way, so that the way flowreplay handles the traffic
will be deterministic. Perhaps, I'll make it easy for others to write
a plug-in which will change it, but that's not something I'm going
to concern myself with now.</SPAN>
<P>
<SPAN ID="hue176">Missing packets in the pcap file will probably make
that flow unplayable. There are proabably certain situation where
we can make an educated guess, but this is far too complex to worry
about for the first stable release.</SPAN>
<P>
<SPAN ID="hue178">That still leaves creating a basic TCP/IP stack
in user space. The good news it that there is already a library which
does this called libnids. As of version 1.17, libnids can process
packets from a pcap savefile (it's not documented in the man page,
but the code is there).</SPAN>
<P>
<SPAN ID="hue180">A potential problem with libnids though is that
it has to maintain it's own state/cache system. This not only means
additional overhead, but jumping around in the pcap file as I'm planning
on doing to handle multiple simultaneous flows is likely to really
confuse libnids' state engine. Also, libnids is licensed under the
GPL, but I want flowreplay released under a BSD-like license; I need
to research if the two are compatible in this way.</SPAN>
<P>
<SPAN ID="hue182">Possible solutions:</SPAN>
<P>

<UL>
<LI><SPAN ID="hue185">Developing a custom wedge between the capture file
and libnids which will cause each packet to only be processed a single
time.</SPAN>
</LI>
<LI><SPAN ID="hue187">Use libnids to process the pcap file into a new
flow-based format, effectively putting the TCP/IP stack into a dedicated
utility.</SPAN>
</LI>
<LI><SPAN ID="hue189">Develop a custom user-space TCP/IP stack, perhaps
based on a BSD TCP/IP stack, much like libnids is based on Linux 2.0.37.</SPAN>
</LI>
<LI><SPAN ID="hue191">Screw it and say that IP fragmentation and out of
order IP packets/TCP segments are not supported. Not sure if this
will meet the needs of potential users.</SPAN>
</LI>
</UL>

<P>

<H2><A NAME="SECTION00042000000000000000">
<SPAN CLASS="arabic">4</SPAN>.<SPAN CLASS="arabic">2</SPAN> <SPAN ID="hue194">Blocking</SPAN></A>
</H2>

<P>
<SPAN ID="hue196">As earlier stated, one of the main goals of this
project is to keep things single threaded to make coding plugins easier.
One caveat of that is that any function which blocks will cause serious
problems.</SPAN>
<P>
<SPAN ID="hue198">There are three major cases where blocking is likely
to occur:</SPAN>
<P>

<OL>
<LI><SPAN ID="hue201">Opening a socket</SPAN>
</LI>
<LI><SPAN ID="hue203">Reading from a socket</SPAN>
</LI>
<LI><SPAN ID="hue205">Writing to a socket</SPAN>
</LI>
</OL>
<SPAN ID="hue208">Reading from sockets in a non-blocking manner is
easy to solve for using poll() or select(). Writing to a socket, or
merely opening a TCP socket via connect() however requires a different
method:</SPAN>
<P>
<BLOCKQUOTE>
<SPAN ID="hue211">It is possible to do non-blocking IO on sockets
by setting the O_NONBLOCK flag on a socket file descriptor using
fcntl(2). Then all operations that would block will (usually) return
with EAGAIN (operation should be retried later); connect(2) will return
EINPROGRESS error. The user can then wait for various events via poll(2)
or select(2).</SPAN><A NAME="tex2html7"
  HREF="#foot383"><SUP><SPAN CLASS="arabic">7</SPAN></SUP></A>
</BLOCKQUOTE>
<SPAN ID="hue216">If connect() returns EINPROGRESS, then we'll just
have to do something like this:</SPAN>
<P>

<DL COMPACT>
<DT>
<DD><SPAN ID="hue219">int&nbsp;e,&nbsp;len=sizeof(e);</SPAN>
<P>
<SPAN ID="hue221">if&nbsp;(getsockopt(conn-&gt;s,&nbsp;SOL_SOCKET,&nbsp;SO_ERROR,&nbsp;&amp;e,&nbsp;&amp;len)&nbsp;&lt;&nbsp;0)&nbsp;{&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue384">&nbsp;&nbsp;&nbsp;/*&nbsp;not&nbsp;yet&nbsp;*/</SPAN>
<P>
&nbsp;<SPAN ID="hue385">&nbsp;&nbsp;&nbsp;if(errno&nbsp;!=&nbsp;EINPROGRESS){&nbsp;&nbsp;/*&nbsp;yuck.&nbsp;kill&nbsp;it.&nbsp;*/&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue386">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;log_fn(LOG_DEBUG,&#34;in-progress&nbsp;connect&nbsp;failed.&nbsp;Removing.&#34;);&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue232">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return&nbsp;-1;&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue234">&nbsp;&nbsp;&nbsp;}&nbsp;else&nbsp;{&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue387">&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;return&nbsp;0;&nbsp;/*&nbsp;no&nbsp;change,&nbsp;see&nbsp;if&nbsp;next&nbsp;time&nbsp;is&nbsp;better&nbsp;*/&nbsp;</SPAN>
<P>
&nbsp;<SPAN ID="hue239">&nbsp;&nbsp;&nbsp;}&nbsp;</SPAN>
<P>
<SPAN ID="hue241">}&nbsp;</SPAN>
<P>
<SPAN ID="hue388">/*&nbsp;the&nbsp;connect&nbsp;has&nbsp;finished.&nbsp;*/&nbsp;</SPAN>
</DD>
</DL><BLOCKQUOTE>
<SPAN ID="hue248">Note: It may not be totally right, but it works
ok. (that chunk of code gets called after poll returns the socket
as writable. if poll returns it as readable, then it's probably because
of eof, connect fails. You must poll for both.</SPAN>
</BLOCKQUOTE>

<P>
<BR><HR><H4>Footnotes</H4>
<DL>
<DT><A NAME="foot383">...&nbsp;</A><A
 HREF="node4.html#tex2html7"><SUP><SPAN CLASS="arabic">7</SPAN></SUP></A></DT>
<DD><SPAN ID="hue213">socket(7)</SPAN>

</DD>
</DL>
<DIV CLASS="navigation"><HR>
<!--Navigation Panel-->
<A NAME="tex2html72"
  HREF="node5.html">
<IMG WIDTH="37" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="next" SRC="next.png"></A> 
<A NAME="tex2html70"
  HREF="flowreplay.html">
<IMG WIDTH="26" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="up" SRC="up.png"></A> 
<A NAME="tex2html64"
  HREF="node3.html">
<IMG WIDTH="63" HEIGHT="24" ALIGN="BOTTOM" BORDER="0" ALT="previous" SRC="prev.png"></A>   
<BR>
<B> Next:</B> <A NAME="tex2html73"
  HREF="node5.html">5 pcap vs flow</A>
<B> Up:</B> <A NAME="tex2html71"
  HREF="flowreplay.html">Flowreplay Design Notes</A>
<B> Previous:</B> <A NAME="tex2html65"
  HREF="node3.html">3 Design Thoughts</A></DIV>
<!--End of Navigation Panel-->
<ADDRESS>
Aaron Turner
2006-07-17
</ADDRESS>
</BODY>
</HTML>