05-06-2018 10:09 PM
New router in, noticed that the sync speed is changing more on this one than the old router, I had to do a few reboots to apply settings and get it working with 2nd router. After each reboot the sync speed changed by a couple of mbps. Its currently sitting at 76071kbps however still only about 69mbps on a speed test due to the error correction I guess.
Looks as though its still on older firmware too as upload attenuation is missing.
05-06-2018 10:38 PM - last edited: 05-06-2018 10:43 PM
I'm on 74.7Mbps throughput on an 80Mbps synchronisation with 10dB noise margin. So 69Mbps with your current synchronisation speed and noise margin is about right. But would expect a few more Mbps when the synchronisation goes back up to establish low error count. Speed measurements with SamKnows Whitebox on my own connection.
Low error count on your line is around 72Mbps with no drops for it is going to run a few more errors than my line. Will ignore 1 line drop for each day currently for DLM has been upset with more than 3 drops in an hour and this we will do for next 3 days.
Is your telephone cable going overhead or completely underground?
Please post a BQM for each day.
05-06-2018 11:02 PM
Okay makes sense, I'd expect at least 74Mbps if the router re-syncs at 80Mbps as this is how it was for a few months but maybe not possible again if line has been affected by others.
Will keep posting the tracker each morning.
Telephone line is completely underground for our street however on the main road it goes overhead so not sure whether it joins those lines overhead or continues underground right to the cabinet and exchange.
05-06-2018 11:15 PM - last edited: 05-06-2018 11:16 PM
It will be underground from the cabinet to the exchange for its fibre cable. If we hit 72Mbps or higher throughput with zero drops over an extended period that is a fix as far as I'm concerned and latency is stable, in my opinion of course.
05-06-2018 11:51 PM
No problem, router stats & speed test attached
|Port||Status||TxPkts||RxPkts||Collision Pkts||Tx b/s||Rx b/s||Up Time|
|WLAN (2.4 GHz)||Disable||0||0||0||0||0||00:00:00|
|WLAN (5 GHz)||Disable||0||0||0||0||0||00:00:00|
|Connection Speed||76071 kbps||19999 kbps|
|Line Attenuation||18.6 dB||0.0 dB|
|Noise Margin||5.8 dB||7.7 dB|
Speedtest results are
To the Structured Communications server in London
06-06-2018 09:43 AM
Attached is the monitor for this morning, no drops on the new router as of yet, the major packet loss last night was when I changed the routers over.
I am noticing though that the ping is quite high only when the network is being used which seems strange to me because when it is in use its not exactly being stretched so theres no reason for a ping reply to spike so much just because someone is watching YouTube or browsing the Internet in the morning...?
06-06-2018 04:27 PM
I've been running a BQM on my own connection to the Sky Q router (ER110) but I don't have Sky Q boxes. It shows that you do get latency jumps during use of the fibre service using YouTube and can conclude from this that the Sky Q routers treat ICMP packets (ping) with very low priority. And that video streaming has peak traffic latency spikes when videos start and it does affect latency measurements (about 3-5ms) at the boundary Sky Q router.
The latency during little use of the fibre service has peak traffic latency occasionally of 1ms but mainly remains on the green line for 80% of the time. This does not agree with my internal latency tests on my own connection with SamKnows Whitebox or, latency tests from an Ubuntu Linux box which is flat so the BQM measurements to the Sky Q router are only of little use for testing latency and using it to detect errors.
It is clear from the previous latency tests done on your Windows machine that it does show issues whether these now show I'm not sure. I'd suggest you do some further latency tests using a *nix machine if you can to get accurate measurements.
06-06-2018 05:33 PM
Did seem a bit strange having the high ping spikes at peak times so its good to see your connection in comparison. Also strange that you would have a light packet loss as well when just streaming? The Sky router must really put ping packets on the low priority as I can see a few dropped packets between 11am & 1:30pm on your BQM.
It's unlikely I'll be able to run any tests on a real Linux box for a while, the only chance I have currently is a virtualized one on Hyper-V which probably wont help much.
What tests do you suggest I run? It may be worth just re-trying them on a Windows machine to see if they differ.
06-06-2018 05:37 PM - last edited: 06-06-2018 05:49 PM
Keep testing on Windows which isn't highly accurate but better than the BQM to the Sky Q router. Even though my use was low from 2:15pm until later it does have 25 or so devices connected to the internet and a website always being used. I'm sure if I had no devices connected it would be flat but that means taking my website offline! I also have a SamKnows Whitebox that kicks in and test the connection whilst not or in low use.