[AusNOG] Why not Symmetric ingress and egress?

Mark Smith nanog at 85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org
Fri Jun 18 08:40:03 EST 2010


Just quickly,

On Thu, 17 Jun 2010 22:02:21 +1000
Paul Brooks <pbrooks-ausnog at layer10.com.au> wrote:

> On 17/06/2010 8:09 PM, Mark Smith wrote:
> >
> > My guess is that ADSL was chosen as the broadband technology to use,
> > rather than a symmetrical DSL technology (not in Australia, I'm
> > talking by the broadband groups who standardise it i.e. the Annex M
> > people), because it sounded right for the way people were using the
> > Internet at the time (consumers rather than producers), rather than
> > understanding that the Internet protocols have operated over symmetric
> > links for most of their life and it is therefore an unstated design
> > assumption. If that wasn't the case, I don't think the above RFC would
> > exist and be a Best Current Practice RFC.
> >    
> Nice try Mark, but your guess would be wrong!. ADSL was developed as a 
> technology for video delivery
                 ^^^^^^^^^^^^^^

Exactly - a purely asymmetric traffic profile. There'll never be a
reason in a video delivery system for a high bandwidth backchannel -
because it is only for video signalling and nothing else.


> several years before Berners-Lee invented 
> HTTP, back when 'broadband' meant using multiple RF carriers over a wide 
> frequency range on coaxial cable, and many years before anyone 
> considered a mass market for Internet connectivity existed, and it was 
> being used for email, gopher, linking university library catalogues, and 
> providing remote terminal access through TELNET protocol to mainframes. 
> ADSL was developed by Bell Labs to allow them to deliver television 
> (over ATM) over their copper wire network, in response to the cable TV 
> network operators starting to do telephone calls, and so competing with 
> the telco's main product.
> 
> ADSL is highly asymmetric because originally that works for TV - high 
> bandwidth towards the viewer, and a bit of capacity for channel-change 
> information back to the network - nothing to do with the Internet. It 
> then started being used for Internet because it matched people's traffic 
> patterns, and also had much longer range than the early HDSL symmetric 
> systems which stopped working after about 2.3km (depending on wire guage).
> 

Extended distance over copper is possibly an early valid reason. It
certainly means that FTTH shouldn't deployed asymmetrically.

> I'd also challenge that TCP/IP has worked and was designed for symmetric 
> links for most of its life. Most of its life, and particularly in the 
> early days when it was being developed, TCP/IP has been carried over 
> serial circuits and dial-up - in the early days 1200/75 bps (asymmetric, 
> deliberately because it matched peoples use of online resources at the 
> time - fast enough to carry keypresses at a moderate typing rate towards 
> the timeshared mainframe, and faster towards the user to provide fast 
> full-screen refresh on 80x25 character terminals and line printers), and 
> later still over 56kbps dial-up, which was actually 56kbps down, 33 kbps 
> upstream assymmetric- or thereabouts. Granted early high-bandwidth 
> trunks between sites were often symmetric 56kbps leased-lines, and later 
> up to 1500 kps T1s until fairly recently when SDH was developed ;-)
> 

I think you're being a bit selective. What about 300/300, 2400/2400,
9600/9600, 14400/14400, 28800/28800, 36600/36000, 64Kbps/64Kbps ISDN,
4Mbps/4Mbps Token Ring, 10Mbps/10Mbps Ethernet, 16Mbps/16Mbps Token
Ring, 100Mbps/100Mbps Ethernet, Symmetric T1s/E1s/J1s, Symmetric Frame
Relay (can it be deployed asymmetrically? possibly, but I've never
heard of it being done that way), 25Mbps Symmetric ATM, 155Mbps
Symmetric ATM, 655 Mbps Symmetric ATM, Symmetric SDH/SONET, CDWM, DWDM
etc. etc. In that historical context (i.e. the last 30 years, not
using popularity as a measure - remember that we all know who Britanny
Spears is, despite there being far better musicians and song writers out
there) of link technologies that have been used to deploy Internet
protocols, asymmetric links, such as 1200/75, cable, ADSL, 56K analog
modems, are exceptions, not the common case. There would be no reason
for the RFC I pointed to to exist if TCP's operation wasn't negatively
impacted by asymmetry.

> Still - lets not allow facts to get in the way of a good story! :-)
> 

I have no reason to make sh** up. Correcting assumptions people make
without any investigation of the facts or history is far more useful,
because then we might actually benefit from the technologies that are
deployed, rather than them being hamstrung by old assumptions, or
making it look like the old version, but faster (sort of like how PPPoE
is like dialup only faster - thankfully the BBF are starting work out
the overheads of that assumption).

(I didn't mention the ADSL video story, but my reference is Computer
Networks 4th edition)

> Regards,
>      Paul.
> 
> 
> _______________________________________________
> AusNOG mailing list
> AusNOG at lists.ausnog.net
> http://lists.ausnog.net/mailman/listinfo/ausnog



More information about the AusNOG mailing list