Odd TBB test ?

[#1]

Very poor performance today. Haven't seen anything this bad until now; started around 5~6pm. Opening sites is close to impossible when my tablet is connected (seems like it wants to max out the limited bandwidth with Windows 10 updates)

Gateway:

Which gateway
You are currently connected to gateway pcn-bng04.

This is located in Colindale Point of Presence.

Traceroute

Tracing route to ntp.plus.net [212.159.13.49]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  LINKSYSHOME [192.168.1.1]
  2    11 ms    11 ms    11 ms  lo0.central10.pcn-bng04.plus.net [195.166.130.25
1]
  3    12 ms    11 ms    11 ms  411.be8.pcn-ir02.plus.net [84.93.253.95]
  4    12 ms    12 ms    12 ms  be1.pcn-ir01.plus.net [195.166.129.160]
  5    12 ms    11 ms    11 ms  cdns01.plus.net [212.159.13.49]

Trace complete.

BT BRAS is currently at 71.12Mbps, tests over wired.

Graph also shows packet loss.

Have also tried rebooting router and manually disconnecting from WAN and cannot seem to "hop" to another gateway. Those old ag gateways were the best for my connection 

Tracing route to bbc.co.uk [212.58.244.22]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  LINKSYSHOME [192.168.1.1]
  2    11 ms    11 ms    11 ms  lo0.central10.pcn-bng04.plus.net [195.166.130.251]
  3    12 ms    12 ms    11 ms  411.be8.pcn-ir01.plus.net [84.93.253.91]
  4    12 ms    12 ms    12 ms  core1-BE1.colindale.ukcore.bt.net [195.99.125.132]
  5    12 ms    33 ms    12 ms  62.172.103.13
  6    39 ms    25 ms    13 ms  194.74.65.42
  7     *        *        *     Request timed out.
  8     *        *        *     Request timed out.
  9    13 ms    12 ms    13 ms  ae0.er01.telhc.bbc.co.uk [132.185.254.109]
 10    13 ms    13 ms    13 ms  132.185.255.148
 11    12 ms    12 ms    12 ms  212.58.244.22

Weirdly enough, traceroutes don't show any high latency but when something is downloading then we can see a high spike to over 200ms.

[#2]

After five or so reconnects, it eventually gateway hopped onto psb-bng04

You are currently connected to gateway psb-bng04.

This is located in Southbank Point of Presence.

It is definitely better, but not as good as how it used to be... I guess we have congestion or something on that previous gateway.

[#3]

Hmmm...

5 minutes later...

You are currently connected to gateway pcn-bng04.

This is located in Colindale Point of Presence.

[#4]

Host	%	Sent	Recv	Best	Avrg	Wrst	Last
plusnet-hub	1	18078	18077	0	0	18	0
lo0.central10.pcn-bng04.plus.net	1	18072	18069	0	8	63	34
411.be8.pcn-ir01.plus.net	1	18070	18067	0	8	60	8
195.99.125.136	1	18071	18068	0	7	123	8
peer1-et-10-3-0.telehouse.ukcore.bt.net	1	18072	18069	0	8	123	8
No response from host	100	3620	0	0	0	0	0
po4-31.core-rs2.thdo.ncuk.net	1	18071	18068	0	15	463	9
pingbox1.thinkbroadband.com	1	18072	18069	0	8	215	8

This is the last 5 hours on PCN-BNG04, one ping drop but that was on my local hub, so no packetloss on that gateway for quite a while.

[#5]

@MrSilver

What does packet loss have to do with the x1 TTB speed test showing 1Mbps speed while the x6 test show 75Mbps? 

I think @Riza has a different issue as those tests show a different pattern. Mine, @rongtw and @Browni show a very distinct pattern of a constant straight line x1 test which is unlike @Riza 's results.

This topic "Odd TBB test" is about these speed tests and mine have never shown any packet loss while reporting these stupid speeds on the x1 test segment. My TBB Quality Monitor is very healthy.

[#6]

Ruling out packet loss has quite a lot to do with it potentially. If single thread is getting packetloss you will take a hit on throughput, one of the charts above talked about potential rwin problems which can manifest through packetloss as the receiver lowers the tcp window size due to loss/retransmits.

When you got to 6 threads some of this can be masked as each of the 6 may be slower but all added up still maxes the line.

So I appreciate you may not have mentioned packetloss and this is not a packetloss thread, but it's important to understand what congestion on a line looks like to help diagnose.

[#7]

Hi

I did a couple of traceroutes today and got:

1 3 ms 3 ms 3 ms dsldevice.lan [192.168.1.254]
2 9 ms 12 ms 13 ms lo0.central10.psb-bng03.plus.net [195.166.130.254]
3 9 ms 10 ms 9 ms 411.be7.psb-ir01.plus.net [84.93.253.115]
4 71 ms 76 ms 73 ms 195.99.125.134
5 22 ms 21 ms 10 ms 213.121.193.207
6 * * * Request timed out.
7 12 ms 9 ms 11 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
8 33 ms 21 ms 13 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
9 10 ms 10 ms 9 ms pingbox1.thinkbroadband.com [80.249.99.164]

1 2 ms 3 ms 3 ms dsldevice.lan [192.168.1.254]
2 9 ms 8 ms 9 ms lo0.central10.psb-bng03.plus.net [195.166.130.254]
3 10 ms 9 ms 9 ms 411.be7.psb-ir01.plus.net [84.93.253.115]
4 10 ms 9 ms 9 ms 195.99.125.134
5 9 ms 10 ms 10 ms 213.121.193.207
6 * * * Request timed out.
7 9 ms 11 ms 9 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
8 21 ms 24 ms 52 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
9 9 ms 9 ms 9 ms pingbox1.thinkbroadband.com [80.249.99.164]

are the above odd? what about the timout in both and the first one has some very long response times

[#8]

You are currently connected to gateway pcn-bng03.

This is located in Colindale Point of Presence.

After a power outage , But now i get back to straight line tests http://www.thinkbroadband.com/speedtest/results.html?id=148190813121144595847

[#9]

@rongtw

Hope it stays that way for you. Mine is still the same as in post 11, wavey lines but in the 70Mbps range (haven't changed gateways since as at least the x1 is in the right ball park)

Haven't had 2 identical straight lines at high speed since I was on an AG gateway.

[#10]

@herojan the timeouts as normal, it's just a device on the route that has been told not to respond to pings.

If the last hop is pinging well spikes in the middle are likely devices that are deprioritising pings in favour of traffic.
Trace routes where the end one is bad are useful to see where it's going wrong.

[#11]

@rongtw I think we must be swapping bad lines lol. Mine is now capping out at 15meg instead of 75! Can I have the good connection back at 9.30 tonight? Lol

It defining looks more like some intermittent issue rather than a core problem but hat probably means longer to find the problem 😳

[#12]

@MrSilver looks like i may have spoken too soon have dropped a bit this morning

http://www.thinkbroadband.com/speedtest/results.html?id=148195403106190485909

[#13]

http://www.thinkbroadband.com/speedtest/results.html?id=1481958461387748655

Mine has got much better this morning!

[#14]

Something is amiss here   another drop

http://www.thinkbroadband.com/speedtest/results.html?id=148195960013732960978

But of course PN will not be able to replicate this

[#15]

Start a protest by downloading flat out 24/7. By using as much bandwidth as possible for as long as possible, you should be making the problem worse so that Plusnet will have to do something about it.