FTTH Speeds

[neiltwist]

I moved into a new build in April (and had to wait a month for OpenReach to turn on the installed line... but that's a different problem) and decided to get the maximum I could, i.e. 300Mbps up, 30Mbps down, since my household is quite a heavy user.

However, I'm finding the whole thing a bit underwhelming. It should be a 300Mbps fibre service, so I would guess should be able to achieve a speed of 250Mbps when plugged directly into the modem and using BT's own speedtest (http://speedtest.btwholesale.com/). However, I currently get as low as 10 Mbps and sometimes the speed drops to virtually nothing (causing streaming and VPNs to drop out). This can happen at any time of day (or night).

My testing setup:

I have tried turning wifi off on the router and used a 1Gbps rated ethernet cable plugged directly into the router.

I have also tried plugging directly into the modem and manually setting up PPPoE.

Finally, I have tried a different router and manually set it up. It is generally a little faster than the HH5, but no surprise there.

Is what I'm doing sensible?

What is a sensible maximum speed that I should be able to get?

What is the best testing setup for max speed?

Thanks all in advance!

[DSLAMdunkin]

Hello Neiltwist

FTTP, of itself, is not a guarantee of high speed as your line is still 'contested' - you still share all of the other fibre infrastructure with other service users.The quoted max speeds are theoretical - you were sensible to have lower expectations than 300Mbs.However, a speed as low as 10Mb would warrant investigation and explanation, by Openreach. Personally, I would expect that yours is a network, rather than a domestic equipment, issue.

I would suspect that you would be entitled to change your service to something more appropriate for your line speed, should it transpire that your issue is congestion related rather than an actual infrastructure fault - why pay for Infinity 4 when your line only supports Infinity 1?...

[pippincp]

---

@DSLAMdunkin wrote:

Hello Neiltwist

FTTP, of itself, is not a guarantee of high speed as your line is still 'contested' - you still share all of the other fibre infrastructure with other service users.The quoted max speeds are theoretical - you were sensible to have lower expectations than 300Mbs.However, a speed as low as 10Mb would warrant investigation and explanation, by Openreach. Personally, I would expect that yours is a network, rather than a domestic equipment, issue.

I would suspect that you would be entitled to change your service to something more appropriate for your line speed, should it transpire that your issue is congestion related rather than an actual infrastructure fault - why pay for Infinity 4 when your line only supports Infinity 1?...

---

You started off good then went awry. FTTP speeds  are not theoretical and the OP should ring in and report it.

[neiltwist]

Thanks for your thoughts DSLAMdunkin, I am quite knowledgeable about networks in general, so I was really looking for some advice on the testing technique and what speeds to expect over FTTP/H.

I've since looked at another post I commented on and found that I should be expecting over 300 Mbps (https://community.bt.com/t5/BT-Infinity-Speed-Connection/FTTP-Infinity-4-330mbps-new-non-Smart-Hub-r...)

Additionally, and thanks for the re-inforcement pippincp, I have already spoke to Retail, who tried to blame it on the ethernet cable I was using, so the purpose of this thread was to work out what the best and most reasonable way of testing the connection.

So, can anyone verify my testing method or suggest a method?

[neiltwist]

Additionally, in moving down from 300 to 250, I took a very conservative view of how much bandwidth I could lose from NAT and a couple of other features of the router.

I wouldn't expect any congestion from BT's own speed test as it should be within the network.

[smf22]

I don't think there's anything flawed in your testing, but here's a couple of things I'd do.

If BT Retail are looking to blame your internal connectivity, the first thing to do is be able to show you don't have a problem there.

If you have multiple wired LAN devices, all with Gigagbit Ethernet capability, then download a program called iperf (https://iperf.fr/iperf-download.php#windows) and test with that. You run it from the command line with one PC as the receiver and the other as the sender. The receiver you enter iperf3 -s, and on the sender you enter iperf3 -c <ip_addr_of_receiver>.

C:\Users\smf22>iperf3 -c 192.168.1.84
Connecting to host 192.168.1.84, port 5201
[  4] local 192.168.1.78 port 11846 connected to 192.168.1.84 port 5201
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-1.00   sec   107 MBytes   894 Mbits/sec
[  4]   1.00-2.00   sec   111 MBytes   930 Mbits/sec
[  4]   2.00-3.00   sec  96.2 MBytes   807 Mbits/sec
[  4]   3.00-4.00   sec   108 MBytes   910 Mbits/sec
[  4]   4.00-5.00   sec   110 MBytes   926 Mbits/sec
[  4]   5.00-6.00   sec   108 MBytes   910 Mbits/sec
[  4]   6.00-7.00   sec   110 MBytes   920 Mbits/sec
[  4]   7.00-8.00   sec  96.6 MBytes   811 Mbits/sec
[  4]   8.00-9.00   sec   111 MBytes   929 Mbits/sec
[  4]   9.00-10.00  sec   109 MBytes   914 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth
[  4]   0.00-10.00  sec  1.04 GBytes   895 Mbits/sec                  sender
[  4]   0.00-10.00  sec  1.04 GBytes   895 Mbits/sec                  receiver

iperf Done.

Depending on the PC specification you should expect 900Mbps or more.

For external testing, the BT Wholesale speedtest servers are the right place to test to as they should be connected to a part the network with high capacity connectivity. So the next thing would be to check connectivity to those servers and make sure you're not seeing packet loss, high 'round trip times' etc. As many servers don't respond to ICMP ping, a tool I tend to use is psping (https://technet.microsoft.com/en-us/sysinternals/psping.aspx) from Microsoft TechNet. One of the tests this can do is to make a TCP connection just as the web based speedtest does, but you can control how many times to connect, the delay in between each connection etc. The following shows 100 (-n 100) connections with no delay (-i 0) between connections. Make sure there is no loss and no spikes in latency.

C:\Users\smf22>psping -i 0 -n 100 speedtest.btwholesale.com:80

PsPing v2.01 - PsPing - ping, latency, bandwidth measurement utility
Copyright (C) 2012-2014 Mark Russinovich
Sysinternals - www.sysinternals.com

TCP connect to 193.113.8.194:80:
101 iterations (warmup 1) connecting test:
Connecting to 193.113.8.194:80 (warmup): 21.17ms
Connecting to 193.113.8.194:80: 19.86ms
Connecting to 193.113.8.194:80: 20.21ms
[snip]
Connecting to 193.113.8.194:80: 21.01ms
Connecting to 193.113.8.194:80: 21.48ms

TCP connect statistics for 193.113.8.194:80:
 Sent = 50, Received = 50, Lost = 0 (0% loss),
 Minimum = 18.84ms, Maximum = 21.88ms, Average = 20.66ms

And then of course you have the speed tests to the Wholesale servers that you've already been running.

The problem could be congestion in the backhaul from your exchange to the core BT sites where the interconnects to the Wholesale servers are. The way to try and demonstrate that is by testing at different times of the day, and avoiding peak times e.g., 18.00 to midnight weekdays.

Regards