Networking question

bill_day at mcgraw-hill.com bill_day at mcgraw-hill.com
Fri Feb 1 12:46:06 CST 2002




Unfortuanatly there is already a group (yes group project) doing IPV6.








Al Tobey <albert.tobey at priority-health.com> on 01/31/2002 10:48:05 PM
                                                                                
                                                                                
                                                                                


                                                              
                                                              
                                                              
 To:      Bill Day/C-I-G at C-I-G                                
                                                              
 cc:      grpm <grand-rapids-pm-list at happyfunball.pm.org>     
                                                              
                                                              
                                                              
 Subject: Re: Networking question                             
                                                              







Hi Bill,

Being the leading Linux bigot on this list, I feel qualified to chime in
here.

   The linux vs. bsd debate has been beat around quite a bit already.
Until kernel 2.4, Linux's network stack was actually based on the BSD
stack.  OS-X (darwin) would make for a nice fresh comparison, but it is
actually based on the BSD network stack with a handful of changes.  What
you'll find is that it's difficult to gauge the efficiency of a network
stack all by itself.   Many things affect the throughput of the stack
such as NIC driver quality, VM, I/O latency, buffering, and point of
focus (security vs. speed vs. scalability).  These may be interesting
things to gauge all on their own.  You'll also notice that there isn't
much to be gained from analysis of the two source codes unless you start
breaking them down into extensive mathematical analyses of various
algorithms within.  For the most part, even this data would be useless
as the two stacks differ quite a bit on how they approach various
problems.  If you really think you can get a lot out of reading the
source, then this would be a fairly original examination as most
comparisons to date are done on the system level with standard
benchmarks and empirical measurements.

What might be more interesting and refreshing is a look at how all of
the IPv6 compliant operating systems are doing things.  This is still
mostly uncovered territory (thus more interesting) and varies a lot
between systems.  This would probably be a bit smaller scope, but would
involve more research and could make way for more background information
(read: fluff) in case you're running short on pages.

-Al

On Thu, 2002-01-31 at 10:54, bill_day at mcgraw-hill.com wrote:
>
>
>
> This is not a Perl question, but I know there are opinionated net savvy people
> on this list.
>
> I'm taking CS 654 (Computer Networking) at GVSU this semester. The class
> requires a survey paper. The potential topics that the professor has suggested
> are really lame. I thought it might be fun to do something like "A comparison
of
> Linux vs. BSD IP implementations, a source code perspective". So I went and
took
> my first look at the Linux kernel source code (net directory). This looks like
> it might be over my head (I'm more of a UNIX wannabe than a guru). On the
other
> hand this is the one time in my life when it may make sense to dive into this
> code.
>
> My questions for the list are:  Is this a reasonable topic? Is my scope too
> large? Does anybody know of good resources other than just the source code? If
> I'm out to lunch, does anybody have ideas for a cool topic for a paper for a
> networking class?
>
> Whatever I end up doing, I'd be willing to present at a future meeting.
>
> Your input is appreciated.
>




********************************************************************
This email and any files transmitted with it are confidential
and intended solely for the use of the individual or entity
to whom they are addressed.  If you have received this
email in error please notify the Priority Health Information
Services Department at (616) 942-0954.
********************************************************************







More information about the grand-rapids-pm-list mailing list