[GRLUG] core 5 iso image?

Bob Kline bob.kline at gmail.com
Sun Apr 30 15:33:22 EDT 2006


That's a somewhat loaded statement I think.  Of course it's the
case that with a clogged server no one does much better than
anyone else.  But even there,  I see no particular reason why
DSL should have an inherent advantage over cable mdem.


I am particularly lucky right now in that in my particular area I rarely
see any loading effect.   I get pretty much the 8Mbps/800Kbps any
time of the day, and any day of the week,  to my hosting service.
But of course that can't apply to servers, which might also have time
of day effects,  or day of the week effects.   DSL has a dedicated
connection from the CO to you,  but that's far as it goes up the line.
A well balanced cable branch should do just as well,  albeit the
cable companies often sell more capacity than they actually have
on a branch,  and then hunker down until enough people scream,
or threaten to move to DSL.  ( Once again the beauty of competition. )


Anyway,  to the extent I am able to test things,  I get the cable bandwidth
I pay for.   I had DSL three years ago,  and it was still expensive for what
I got,  which is why I switched.  I'm sure the bang for the buck ration has
improved a lot since then.  But I still can't see any technical reason why
DSL should perform better per unit of bandwidth over a broad range of
situations.   And in my experience it doesn't.  I have done things like
tweak input buffer sizes,   trying to optimize transfer rates,  and settled
on
32K.   It's not a sharp peak, but can reduce packet overhead a little bit.
But this has little to do with DSL versus cable modem.


The only big effect I've ever seen is between windoz and Linux.  Linux
simply performs better - maybe 10%.  I attribute that simply to Linux being
a better implementation of the protocol stack.  But again has little to do
with DSL versus cable modem.


If anyone out there has any measured numbers on this I would be
interested to see what I am missing.  But having been in both camps,
the only thing that would get me away from the speed of cable modem
would be an overwhelming  cost/performance advantage for DSL.  As much
as I use my computers each day,  bandwidth and performance are worth
a lot to me,  and it will be a tough sell.   DSL could improve it's speed a
lot of the telcos want to spend the money for system upgrades,  but I'll
wait until I see it.  For now the speed is worth more than the cost
difference.


     -Bob


On 4/30/06, Joshua Burns <joshuadburns at hotmail.com> wrote:
>
> Good question. In fact I just spent the last hour trying to back up what I
> said with a web site or two but can't find anything. The only place I can
> find that information is in a book of mine from ITT Technical Institute,
> "Network Standards and Protocols," which states Cable only supports around
> 60 concurrent connections or so, while DSL is generally a couple hundred.
> We'll keep in mijnd however that these books have been published by NIIT
> and
> therefore have absolutely no firm standing on factual information.
>
> I do however stand by my statement that even though my friend can run a
> bandwidth test and get around a 7.5 mb connection, I can still generally
> download files via peer to peer networks close to the same speed he can.
>
> Josh
>
>
> >From: "Tim Schmidt" <timschmidt at gmail.com>
> >Reply-To: grlug at grlug.org
> >To: grlug at grlug.org
> >Subject: Re: [GRLUG] core 5 iso image?
> >Date: Sun, 30 Apr 2006 02:06:25 -0400
> >
> >On 4/30/06, Joshua Burns <joshuadburns at hotmail.com> wrote:
> > > Well something you also must consider is whether or not you're using
> >cable.
> > > If you are, keep in mind you are only allowed so a low number of
> > > simultaneous connections while with T1/3 or DSL, this number is much
> >larger.
> > > Being that peer to peer networks require many, many connections, using
> >cable
> > > will be a sure down fall.
> > >
> > > I have a friend who has 8 Megabit cable while I myself have 1.5 DSL. I
> >still
> > > download things nearly as fast as he does, simply because if you were
> to
> > > monitor the amount of concurrent connections, you'd see my connection
> >has a
> > > lot more actively downloading, and less being choked, while his is the
> > > opposite.
> >
> >What's the reasoning behind this?  As far as I understand, the number
> >of concurrent connections you can handle has everything to do with the
> >computer making the connections, and possibly the routers in between
> >the connections (including your personal router if you have one, and
> >the ISPs), but absolutely nothing to do with the medium carrying the
> >packets.  So explain why cable would be any different than DSL of a
> >similar throughput?
> >
> >--tim
> >_______________________________________________
> >grlug mailing list
> >grlug at grlug.org
> >http://shinobu.grlug.org/cgi-bin/mailman/listinfo/grlug
>
> _________________________________________________________________
> Is your PC infected? Get a FREE online computer virus scan from McAfee(r)
> Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963
>
>
>
> _______________________________________________
> grlug mailing list
> grlug at grlug.org
> http://shinobu.grlug.org/cgi-bin/mailman/listinfo/grlug
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://shinobu.grlug.org/pipermail/grlug/attachments/20060430/f47c92e5/attachment-0001.htm 


More information about the grlug mailing list