Kitz Forum
Internet => General Internet => Topic started by: Weaver on October 14, 2019, 02:06:50 AM
-
Could someone tell me if these calculations are vaguely right
If I assume a distance from Skye to London of ~950 km, the speed of light is 299792458 ms-1 and the refractive index of glass is 1.44, then I make that an RTT = 950000 / 299792458 * 2 / 1.44 = ~9ms.
So I think my measured RTT to AA of 43 ms is 34ms due to interleave plus the distance-related time delay. That figure is from pings to bottomless.aa.net.uk. I thought the delay due to large interleave values was more than that?
-
I dont know for each specific interleave settings, but the default ECI VDSL profile was adding around 8-9ms delay to my line. It is a fixed amount of delay not a %, so a higher base latency has a lower % impact.
-
It's nowhere near as simple as your calculation.
Every bit of kit along the route can add latency.
On OpenReach FTTC/P inner London connections often see around 5ms latency.
Northern parts of England tends to be in the 10-15ms range.
Central Scotland is around 15-20ms.
An OpenReach FTTP connection in Inverness is about 22ms? if I recall correctly.
9ms from Skye to London might be theoretically possible but not gonna happen on consumer kit with dozens of points inbetween.
Does the "adsl info --stats" command not provide a delay figure on ADSL?
FTTC low interleaving adds exactly 8ms.
-
@j0hn I wondered about the per hop additional delays. But they aren’t distance related and I was just, for the moment, interested in the effect of geography. Clearly though, more distance means a good chance of more hops.
-
FYI when I was on 20CN ADSL my RTT was around 25ms on fastpath, 45ms when interleaved. That was RTT from a host on my LAN to ntp.plus.net.
-
In reply #2, above, j0hn asked --
Does the "adsl info --stats" command not provide a delay figure on ADSL?
A delay figure is, indeed, shown for an ADSL2 connection. What follows is harvested from my own circuit, with a pruning of unnecessary lines from the beginning and from the end --
<snip>
Link Power State: L0
Mode: ADSL2 Annex A
TPS-TC: ATM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.1 6.2
Attn(dB): 45.5 27.4
Pwr(dBm): 0.0 12.7
ADSL2 framing
Bearer 0
MSGc: 59 12
B: 182 127
M: 1 1
T: 1 1
R: 14 10
S: 1.0000 3.9856
L: 1576 277
D: 64 8
Counters
Bearer 0
SF: 400699 363085
SFErr: 36 0
RS: 26045368 2235999
RSCorr: 56667 0
RSUnCorr: 960 0
Bearer 0
HEC: 617 0
OCD: 2 0
LCD: 2 0
Total Cells: 89438218 15628429
Data Cells: 1243500 122197
Drop Cells: 0
Bit Errors: 40386 0
ES: 9 0
SES: 0 0
UAS: 46 46
AS: 6513
Bearer 0
INP: 2.00 1.00
INPRein: 0.00 0.00
delay: 16 8
PER: 16.25 17.93
OR: 32.00 8.02
AgR: 5833.21 1023.71
Bitswap: 1439/1439 0/0
<snip>
-
16ms on the downstream and 8ms on the upstream.
Would Weaver need an identical delay on each line or would the Firebrick deal with the difference?
Having PhyR activated I'd be tempted to try without the interleaving/delay if possible as that should immediately reduce the latency by whatever the delay figure is.
Not something I'd try on a 7km ADSL line without PhyR.
-
I suspect attempting to switch off the downstream interleaving may switch off the downstream PhyR. It's currently doing PhyR instead of standard FEC+interleaving on the downstream. It may well be using some FEC with the PhyR, but they'll be no interleaving of any significance.
-
Completely missed this thread. Too old to be hitting the tiles so catching up.
Modulating and demodulating the DSL isn't immediate.
The fibre route will be circuitous - networks are not optimised for latency unless manually tuned or built for it.
The intermediate devices will add latency of varying degrees depending on the processing and configuration they are doing.
Etc.
The lowest latency links are not over fibre. The microwave dishes on top of financial houses are there for good reason :)
-
Microwave links are low-latency ?
-
Yup. Due to refractive index across fibre the signal travels at about 2/3rds the speed of light, microwave transmission runs at about 99% of the speed of light.
https://arstechnica.com/information-technology/2016/11/private-microwave-networks-financial-hft/
https://en.wikipedia.org/wiki/High-frequency_trading#Low-latency_strategies
Let's take London-Frankfurt as a real-world example. As the raven flies, the two cities are 396 miles apart. A radio signal travelling through air at just under the speed of light (299,700km per second) would cover that distance in 2.126 milliseconds. Through a glass or plastic fibre, where light has to bounce along the refractive index rather than travel in a straight line, the speed of light is reduced to around 200,000km/s, resulting in a theoretical minimum latency of 3.186 milliseconds.
In reality though, the fibre network between London and Frankfurt isn't just a single straight piece of fibre. For a start, depending on where the London server is, the packet of data might bounce around a few times until it gets to the right router for its journey across to Europe. Along the way, there are other routers and repeaters. And once the packet arrives in Frankfurt, it's the same deal as in London: the destination server is probably a few hops away.
Add geography and infrastructure to the mix, a submarine cable crossing (probably via Calais in France or Ostend in Belgium), plus the fact that a router in London might decide to send the packet via Paris instead, and the average latency between two servers in London and Frankfurt is actually closer to 17 milliseconds.
Now, at long last, for the punchline: a private microwave network between London and Frankfurt has a latency of about 4.2 milliseconds. I say "about," because the newest connections are still secret and their exact latencies are unknown. That's why businesses and financial institutions which absolutely must have the fastest connection opt for microwave links instead of fibre.
-
Ah of course. I was thinking about the nodes only, not the links, for the moment. I had done the refractive index calculation a while back but I have no idea about the additional delay time contributions at each node in the graph.
-
Very, very low using decent kit. No doubt these finance houses are using more than just decent kit.
-
Very, very low using decent kit. No doubt these finance houses are using more than just decent kit.
You mean I can't get this on Ubiquiti? I feel ripped off. :p
-
Pop you head right next to a very long-range microwave link dish, in between it and a peer, and contemplate it. You'll feel better in no time. ;)
-
In my youth I used to marvel at these microwave towers. From the Morridge high viewpoint of the Staffordshire Moorlands I could see the tower on Cannock Chase at Pye Green to the south and the one at Sutton Common (vaguely) near Macclesfield in Cheshire to the northwest.
-
Indeed! Before fibre optics and after coaxial those were really important to the UK's telecomms network.