topic Re: Line speed v Sync speed in Plusnet Feedback

Line speed v Sync speed

glloyd — Mon, 02 Jul 2012 11:16:56 GMT

Why is the line speed so much lower than the sync speed. It can't be all overheads otherwise some people on lower connect speeds would not get any throughput at all. It seems to me that there is a certain amount of throttling.

Re: Line speed v Sync speed

Anotherone — Mon, 02 Jul 2012 12:57:11 GMT

As you ought to know "Line speed" (if you are talking about that at the member centre) is the equivalent of the BT IP profile which on 21CN is 88.2% of the sync speed. On 20CN the profile is defined in bands by the table that's been published for years - I thought you knew all this. If not you should pop over to kitz and have a good read.

Re: Line speed v Sync speed

glloyd — Fri, 06 Jul 2012 11:57:27 GMT

Is there any need to be so damn rude. Of course I know the IP profile is about 88% of sync speed and the reason given is overheads. BUT, and the reason for my post, that is a big chunk of the sync speed. Surely there could be a way of cutting those overheads down, I don't think they were always that high.

Re: Line speed v Sync speed

Oldjim — Fri, 06 Jul 2012 12:16:43 GMT

Interesting question
A quick search threw up this http://www.dslreports.com/speed/

Quote
In an ideal world, you should be able to see in your browser download window, during a sustained transfer, a rate equal to your purchased speed, divided by 8 (to get bytes), less 13% (TCP/IP and ATM header overhead).

another search found this nugget

Quote
ATM cells are 53 bytes long: 5 header and 48 payload

I wonder if the ATM overhead is 5/48 or 10.4%
This is a much better explanation http://aa.net.uk/kb-broadband-how-atm.html

Re: Line speed v Sync speed

spraxyt — Fri, 06 Jul 2012 13:32:41 GMT

Yes, the bulk of the 11.8% overhead is from ATM. As mentioned ATM cells are 53bytes long with 5byte header and 48byte payload. That makes the ATM overhead 5/53 * 100% = 9.4%. One of the reasons ATM packets are so small is because they are designed to carry different types of information - audio as well as TCP/IP data. Note this is telephone type audio, not VoIP which uses TCP/IP for transport.
TCP/IP overhead is 40bytes so 40/1500 * 100% = 2.7% in a 1500byte packet. PPP also incurs a small overhead. There is also wastage because I don't think an ATM cell is permitted to carry data from more than one TCP/IP packet.
In fixed speed ADSL services these overheads were still present but "built in". So, for example with a 512kbps (0.5Mbps) service the sync speed would actually be 576kbps; a 1Mbps service would sync at 1152kbps; and a 2Mbps service would sync at 2272kbps.

Re: Line speed v Sync speed

glloyd — Sat, 07 Jul 2012 10:09:42 GMT

I wonder if the full fibre exchanges i.e. no copper make any difference to overheads?

Re: Line speed v Sync speed

Oldjim — Sat, 07 Jul 2012 10:13:59 GMT

It won't as the transmission protocols will remain the same

Re: Line speed v Sync speed

spraxyt — Sat, 07 Jul 2012 10:47:25 GMT

I think fibre might make a difference since I believe end-to-end PPPoE is used (ie no ATM). I've seen mention in the Fibre Trials board that the overhead on FTTC is only 3% which looks like just TCP/IP and PPP.
FTTC knows conventional analogue voice is being used so ATM audio needn't be allowed for. I assume FTTP operates similarly at the moment, with "no copper" to come (on BT services).

Re: Line speed v Sync speed

Anotherone — Sun, 08 Jul 2012 22:09:24 GMT

Quote from: glloyd
Is there any need to be so damn rude. Of course I know the IP profile is about 88% of sync speed and the reason given is overheads. BUT, and the reason for my post.........

As you obviously knew that, then if your question had been phrased differently and asked in the right place - it's not feedback on Plusnet's service, then you wouldn't have got such a sarcastic response. Plusnet are, as is any ISP, only operating a system to Internationally agreed standards, and you won't change that! (as frustrating as it maybe, nor do I disagree with your sentiment).

Re: Line speed v Sync speed

Oldjim — Mon, 09 Jul 2012 08:09:45 GMT

jim:post Perhaps if posters have nothing positive to contribute and only want to make a sarcastic comment then they should refrain from responding mod:end

Re: Line speed v Sync speed

Anotherone — Mon, 09 Jul 2012 09:11:50 GMT

If that remark is aimed at me Jim, then I suggest you re-read my initial response where I did point the OP to a web site that would have led to plenty of information on the issue. And this board is still the wrong board for this topic IMHO

Re: Line speed v Sync speed

Oldjim — Mon, 09 Jul 2012 09:16:59 GMT

if the cap fits
If it was thought to be the wrong board a Report to Moderator would have been appropriate and I am not going to move it now

Re: Line speed v Sync speed

Anotherone — Mon, 09 Jul 2012 09:27:14 GMT

I did a report to moderator

Re: Line speed v Sync speed

Oldjim — Mon, 09 Jul 2012 09:56:05 GMT

5 days after your original sarcastic response

Re: Line speed v Sync speed

Anotherone — Mon, 09 Jul 2012 10:08:45 GMT

Jim - you, Dick and Adie are normally quite on the ball when it comes to moving posts. If I were to stop and report every thread I thought was on the wrong board everytime I spotted one, it would waste a fair bit of my time, never mind yours as your moderation boards would get cluttered with stuff you'd have to read when you may have already dealt with the matter by the time your read it.
However your comments are noted - I will refrain in future from posting at all in any threads I consider are on the wrong board and I shall click on report to moderator.