Awful web browser download speeds but torrents are fine?

Hi folks

Wonder if you can help shed some light on to my problem.

I've had Standard BT Broadband (fibre not available yet) for 6 months in a brand new apartment block.

The connection is stable. I connect at 6Mbps which isn't great but expected. My download speeds prior to this problem usually max out around 650 KB/s.

However, my download speeds have since plummeted to 0-40 KB/s when downloading from a website, web browsing, streaming is impossible. It's pretty much unusable.

I have run various speed tests, at different times in the day, on different devices, on the master socket, extension sockets, switch microfilters, even tried the test master socket. Nothing is working.

This same problem happened a few months back and BT sent me a new Home Hub 4 which did return the performance back to normal for a few weeks. But it's back again.

The weird thing is, BItTorrent download speeds are unaffected and I still can max out at 650KB/s with them. How can this be?

I have been on the phone/live chat to BT, they have run various tests and made changes but nothing has worked.

Any help or advice greatly appreciated as this is really frustrating!

Are you acting as a BitTorrent seed? Maxing out your upstream connection by acting as a seed can badly affect your downstream performance.

No - I only run uTorrent as a test. I close it down after that - it's not running in the background.

What figures does the ThinkBroadband speed tester give?

More importantly post the link to the page with the graphs on.

http://www.thinkbroadband.com/speedtest/button/14713...

When you say you've been on the phone to them, was that using the landline, or your mobile? Is there a dial tone on the landline?

In reply to a post by RobertoS:

When you say you've been on the phone to them, was that using the landline, or your mobile? Is there a dial tone on the landline?

Mobile - I don't have a phone for the landline.

BT just called me to say the fault has been resolved.

Sure...

Test after I got off the phone to them, and also after a router restart:

http://www.thinkbroadband.com/speedtest/button/14713...

Is that with a wired connection to the router or wireless? Also, what are the router stats?

In reply to a post by micksharpe:

Is that with a wired connection to the router or wireless? Also, what are the router stats?

I'm performing these tests over ethernet, but the slow speeds are also apparent on my wireless devices - Macbook, mobile phone, Fire Stick

Latency is pretty horrendous. Maybe you have a DNS/routing problem. Can you do a traceroute? Say:

tracert thinkbroadband.com

Edited by micksharpe (Tue 16-Aug-16 18:08:00)

Pinging 8.8.8.8 is a steady 13ms. 0% packet loss.



Tracing route to thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 11 ms 10 ms 10 ms 31.55.186.184
5 11 ms 11 ms 11 ms 195.99.127.58
6 12 ms 12 ms 12 ms host213-121-193-165.ukcore.bt.net [213.121.193.165]
7 11 ms 11 ms 10 ms linx-gw2.thdo.ncuk.net [195.66.236.240]
8 26 ms 27 ms 12 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
9 11 ms 10 ms 11 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.

That sort of behaviour of the broadband can often be caused by one wire of the line pair being disconnected at some point. Are you able to get the line stats (data rate/connection speed, attenuation and noise margin) from the router? I don't think you can with an HH4.

Anyone with landline broadband needs a cheap corded phone such as this one for checking for a disconnected wire. Or for a noisy line. Even if you never use the phone for anything else and even if in this case that isn't the cause. Mine is stuffed away in a junk room cupboard.

£10 is a lot for one. here's a cheaper one if you are near an Argos.

With only one wire connected broadband can connect at a drastically reduced speed. A landline phone won't work at all - just dead. A corded rather than a cordless one is needed when checking for simple noisy/crackly lines. Which may have been how your problem started, with a loose wire. The router replacement working could just have been coincidence.

In reply to a post by RobertoS:

£10 is a lot for one. here's a cheaper one if you are near an Argos.

This is even cheaper if you can get to Argos.

But OOPS!

I've just realised the OP says BitTorrent is working fine. Which almost certainly scuppers my entire theory.

They still should have a corded phone, for what goes wrong tomorrow .

Does the problem persist if you disable wireless on the router?

In reply to a post by micksharpe:

Does the problem persist if you disable wireless on the router?

Yep I tested that the other day. Still persists. I've just disabled both 2.4 GHz and 5 GHz again and done some more testing.

So the only thing connected to the Home Hub right now is my Windows 10 PC over ethernet.

Current ADSL line status is (this is expected)
Downstream: 6.398 Mbps
Upstream: 1.062 Mbps

Trace route
Tracing route to thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 11 ms 10 ms 10 ms 31.55.186.184
5 11 ms 11 ms 11 ms 195.99.127.58
6 12 ms 12 ms 12 ms host213-121-193-165.ukcore.bt.net [213.121.193.165]
7 11 ms 10 ms 11 ms linx-gw2.thdo.ncuk.net [195.66.236.240]
8 11 ms 11 ms 11 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
9 11 ms 11 ms 11 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.

Torrent download speeds around 600 KB/s:
https://www.dropbox.com/s/vp1s9r3wjzpghd0/torrent.pn...


Web browser download speeds sits around 50 KB/s (http://download.thinkbroadband.com/1GB.zip)
https://www.dropbox.com/s/snsmpmzhxgvrpun/chrome.png...

Latest http://www.thinkbroadband.com/speedtest.html result:
http://www.thinkbroadband.com/speedtest/button/14713...
The test results seem to be fluctuating up and down a fair bit...


I just can't work this one out. As mentioned earlier, this problem raised its head a couple of months ago and BT sent me a new HH4 which did fix the performance. I am wondering if the HH4 is at fault again....

In reply to a post by David_W:

This is even cheaper if you can get to Argos.

Purchased 47 times in the last 48 hours!

Can you test with another laptop / desktop over wired?

At those slow speeds it is unlikely BT will not be able to handle the load at their end.

In reply to a post by dwjoll:

Pinging 8.8.8.8 is a steady 13ms. 0% packet loss.

Something like -

ping -l 1400 -n 100 8.8.8.8

(where l = lowercase L) would be a better test for packet loss.

It sounds like it might be a good idea to ask BT to send you another Home Hub. At least it's worth trying, especially if it (temporarily) solved the problem last time. Personally, I wouldn't want a modem/router that's been rigged to not display line stats (as current Home Hubs are). You might want to consider investing in a better modem/router.

In reply to a post by legume:

In reply to a post by dwjoll:

Pinging 8.8.8.8 is a steady 13ms. 0% packet loss.

Something like -

ping -l 1400 -n 100 8.8.8.8

(where l = lowercase L) would be a better test for packet loss.

Ping statistics for 8.8.8.8:
Packets: Sent = 100, Received = 80, Lost = 20 (20% loss),
Approximate round trip times in milli-seconds:
Minimum = 27ms, Maximum = 28ms, Average = 27ms

In reply to a post by ukhardy07:

Can you test with another laptop / desktop over wired?

At those slow speeds it is unlikely BT will not be able to handle the load at their end.

Hmm I don't have anything else that is wired.

In reply to a post by micksharpe:

It sounds like it might be a good idea to ask BT to send you another Home Hub. At least it's worth trying, especially if it (temporarily) solved the problem last time. Personally, I wouldn't want a modem/router that's been rigged to not display line stats (as current Home Hubs are). You might want to consider investing in a better modem/router.

Thanks for the advice Mick. I'll call them back tomorrow and ask for a replacement hub, or that they send an engineer out.

The thing I still don't get is the stable torrent download speeds. If there's anyone reading this who can explain the difference why this may be I'd be interested to hear it.

You might want to try running the Windows Resource Monitor to see what processes are using the network. It just might be something to do with Windows rather than the network or your ADSL connection.

The torrents speeds - its easy.

See how on the speed tests we do two download tests, the first is really slow and this is the same as you downloading ONE file at a time, the second that is better is 6 files at once.

Torrents use many files at once, 32, 64, 128 or maybe more and thus can fight through whatever is slowing you down.

The conclusion is that the ADSL2+ itself is fine, but something is either downloading or unaware that you are unaware of on your connection. Another HH4 has gone bad - and I'd go and spend £20 to £30 on my own ADSL2+ modem/router to get away from the BT firmware updates. Some really bad local congestion on the telephone exchange and the IP network to the wider world, in which case try at 6 or 7am (late night testing no good due to Olympics at night)

In reply to a post by dwjoll:

Ping statistics for 8.8.8.8:
Packets: Sent = 100, Received = 80, Lost = 20 (20% loss),
Approximate round trip times in milli-seconds:
Minimum = 27ms, Maximum = 28ms, Average = 27ms

Nasty, hopefully getting another modem/router will help - though if it continues after that, having one where you can see the line stats would be a lot better than some locked down BT thing.

Edit: The steady latency indicates that nothing was flooding upload/download at the time of the test.

Edited by deleted (Wed 17-Aug-16 14:58:06)

Many thanks all for the replies and information.

BT have contacted me today and they are sending an engineer out to me on Monday. I hope they find something.

Can anyone recommend me a steady router? I don't mind spending around £50.

In reply to a post by legume:

In reply to a post by dwjoll:

Ping statistics for 8.8.8.8:
Packets: Sent = 100, Received = 80, Lost = 20 (20% loss),
Approximate round trip times in milli-seconds:
Minimum = 27ms, Maximum = 28ms, Average = 27ms

Nasty, hopefully getting another modem/router will help - though if it continues after that, having one where you can see the line stats would be a lot better than some locked down BT thing.

Edit: The steady latency indicates that nothing was flooding upload/download at the time of the test.

I take it this is bad? Is there an accepted % packet loss for standard broadband?

Yes. Zero.

In reply to a post by RobertoS:

Yes. Zero.

That's a bit misleading. We may all be used to zero packet loss and have therefore come to expect it but the TCP/IP protocol was designed to cope with lost packets. The mantra is "reliable connection over unreliable networks". 20% packet loss when pinging a well-known DNS server is unacceptable, however.

I agree. It was a bit tongue in cheek, but give them a figure . People seem to get very excited if they get intermittent 1% on a tbb BQM or f8lure.

I just cannot believe BT is at 20% packet loss and this bad in your area. I feel it must be the router or the computer in question.

BT are putting out 300Mbps+ connections in plenty of my friends locations, not saying it is not congestion or a network issue, and gosh that speedtest looks like dire congestion, but I am yet to believe it is the case.

Have you still got the old HH4? BT do not usually force them to be returned.

EDIT: could easily be a faulty ethernet card or messed up OS.

Edited by ukhardy07 (Wed 17-Aug-16 17:14:00)

In reply to a post by RobertoS:

but give them a figure...

Hmmm... I can't see BT promising zero packet loss any time soon (even if it's just over their own network).

In reply to a post by dwjoll:

I take it this is bad? Is there an accepted % packet loss for standard broadband?

For that test to a known "good" server I would say usually zero. Of course you are somewhat limited with what you can infer when pinging one server = it could be them, but I get zero several times to 8.8.8.8. Multiple tests to different servers at different times would establish better what's really going on.

As for acceptable for ADSL: IIRC SNRM of 0dB is defined as a BER of 1/10,000,000 which (unless I messed up calculating) is actually not that good = with full size packets 1/737, Hmm. That's (IIRC) 0dB - connections usually run at 6dB so I have no idea what the BER is expected to be for that.

Can you post results of TBB speed test using your Mac (doesn't matter that over WiFi)? And the Terminal command "ping -s 1400 -c 100" for both 8.8.8.8 and your router? I assume you have already tried a factory reset of the HomeHub.

In reply to a post by ukhardy07:

I just cannot believe BT is at 20% packet loss and this bad in your area. I feel it must be the router or the computer in question.

Yea, or maybe some faulty exchange card/port/something

Years ago when ADSL max came out my exchange developed a strange fault where line stats would show no crc errors but aal5 sar stats were showing loss, the inference being atm cells were being lost somewhere at/beyond the exchange. It soon go fixed, searching at the time came up with a description of "exchange out of sync". Though I guess this was when atm went deeper into network than now with 21cn.

In reply to a post by ukhardy07:

EDIT: could easily be a faulty ethernet card or messed up OS.

I was going to suggest re-plugging etc. but I think somewhere the OP said it was the same over wireless.

Edit: Thinking more, if home hubs are known to answer pings lan side reliably then pinging the router would be an interesting test to do.

Edited by deleted (Wed 17-Aug-16 18:02:41)

In reply to a post by ukhardy07:

I just cannot believe BT is at 20% packet loss and this bad in your area. I feel it must be the router or the computer in question.

BT are putting out 300Mbps+ connections in plenty of my friends locations, not saying it is not congestion or a network issue, and gosh that speedtest looks like dire congestion, but I am yet to believe it is the case.

Have you still got the old HH4? BT do not usually force them to be returned.

EDIT: could easily be a faulty ethernet card or messed up OS.

Can't be my computer or network card. Same results on my Macbook pro. I sent the old HH4 back for recycling - doh.

In reply to a post by RobertoS:

People seem to get very excited if they get intermittent 1% on a tbb BQM or f8lure.

Quite right, assuming the line is not flooded then getting 1% loss on one tiny packet that's only being sent once a second is not good at all.

In reply to a post by vimto_girl:

Can you post results of TBB speed test using your Mac (doesn't matter that over WiFi)? And the Terminal command "ping -s 1400 -c 100" for both 8.8.8.8 and your router? I assume you have already tried a factory reset of the HomeHub.

On PC over ethernet (for comparison)
Speedtest
http://www.thinkbroadband.com/speedtest/button/14714...

Ping to 8.8.8.8
Ping statistics for 8.8.8.8:
Packets: Sent = 100, Received = 71, Lost = 29 (29% loss),
Approximate round trip times in milli-seconds:
Minimum = 26ms, Maximum = 27ms, Average = 26ms

Ping to router 192.168.1.254
Ping statistics for 192.168.1.254:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 1ms, Maximum = 1ms, Average = 1ms

On Macbook over wifi
Speedtest
http://www.thinkbroadband.com/speedtest/button/14714...

Ping 8.8.8.8
--- 8.8.8.8 ping statistics ---

100 packets transmitted, 82 packets received, 18.0% packet loss

round-trip min/avg/max/stddev = 26.884/30.763/31.659/0.887 ms


Ping router 192.168.1.254
-- 192.168.1.254 ping statistics ---

100 packets transmitted, 99 packets received, 1.0% packet loss

round-trip min/avg/max/stddev = 3.091/5.742/6.607/0.567 ms

Seems to be more packet loss on PC over ethernet but to be honest I think that's just pot luck as there are times when the performance all of a sudden improves - seemingly at random.

Yes I have tried a factory reset on the router.

Using ping -l 1400 -n 100 8.8.8.8, connected wirelessly to my router which goes through an HG612.

Ping statistics for 8.8.8.8:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 24ms, Maximum = 39ms, Average = 25ms

pinging www.thinkbroadband.com is a good one, since we can cross check our actual network and the BQM from many providers to see if there is a provider wide issue, i.e. insight not available from picking random box on le internet

If its a dead home hub, I'd take a free replacement and then worry about the upgrading and putting the HomeHub away in a cupboard for dealing with any fault handling.

In reply to a post by legume:

Years ago when ADSL max came out my exchange developed a strange fault where line stats would show no crc errors but aal5 sar stats were showing loss, the inference being atm cells were being lost somewhere at/beyond the exchange. It soon go fixed, searching at the time came up with a description of "exchange out of sync". Though I guess this was when atm went deeper into network than now with 21cn.

Unless I'm very much mistaken, 21CN is an MPLS network, with no ATM links. DSL users can use PPPoA if they wish (arguably it's a better option on ADSL than PPPoE, especially on sub-10 Mbit/s links), but that's the only place AAL5 is used. The traffic from the MSAN onwards is MPLS.

This might be a hardware/config issue on the local BT network side. If you have BTWiFi/Fon enabled can you connect to see what results you get? Do you have a new Hub coming in the post or have they left it to the engineer to replace?

In reply to a post by David_W:

In reply to a post by legume:

Years ago when ADSL max came out my exchange developed a strange fault where line stats would show no crc errors but aal5 sar stats were showing loss, the inference being atm cells were being lost somewhere at/beyond the exchange. It soon go fixed, searching at the time came up with a description of "exchange out of sync". Though I guess this was when atm went deeper into network than now with 21cn.

Unless I'm very much mistaken, 21CN is an MPLS network, with no ATM links. DSL users can use PPPoA if they wish (arguably it's a better option on ADSL than PPPoE, especially on sub-10 Mbit/s links), but that's the only place AAL5 is used. The traffic from the MSAN onwards is MPLS.

Yea, AFAIK the only ATM is on the wire now, which is what my last sentence above was supposed to mean - though It wasn't very clear.

I think PPPoE on ADSL is really PPPoEoA so still using AAL5 but with more overhead than PPPoA, which means it's still the best option (unless you can do IPoA). I think TR069/whatever will choose it first.

IIRC from many years ago it was said that BTW only enabled PPPoE(oA) on DSL because AOL lobbied them as they had loads of kit from the US that couldn't do PPPoA. Don't know if it's true or not, but the fact AOL were "players" shows how long ago it was

In reply to a post by MrSaffron:

pinging www.thinkbroadband.com is a good one, since we can cross check our actual network and the BQM from many providers to see if there is a provider wide issue, i.e. insight not available from picking random box on le internet

ping -l 1400 -n 100 www.thinkbroadband.com

Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 80, Lost = 20 (20% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 26ms, Average = 25ms

In reply to a post by vimto_girl:

This might be a hardware/config issue on the local BT network side. If you have BTWiFi/Fon enabled can you connect to see what results you get? Do you have a new Hub coming in the post or have they left it to the engineer to replace?

Yes I have that enabled. Just connected to it and ran a speedtest on my Samsung Galaxy S5. Slightly improved results.
http://www.thinkbroadband.com/speedtest/button/14714...

They aren't sending me a new home hub as far as I know, I guess that depends on what the engineer finds. Last time they sent me a new HH4 straight away and that fixed it for a while.

Edited by deleted (Wed 17-Aug-16 19:49:19)

Okay now try this

ping -l 1400 -n 100 31.55.186.184

This is the IP address from the first hop on a traceroute you posted earlier I believe.

Alas due to daft home hubs we cannot do any pings or traceroutes to your home hub as a reverse trace can be useful.

You can ping a device in the DMZ of the homehub

In reply to a post by MrSaffron:

Okay now try this

ping -l 1400 -n 100 31.55.186.184

This is the IP address from the first hop on a traceroute you posted earlier I believe.

Alas due to daft home hubs we cannot do any pings or traceroutes to your home hub as a reverse trace can be useful.

Can't ping that IP at all.

Just did a fresh tracert:

Tracing route to www.thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 12 ms 12 ms 11 ms 31.55.186.180
5 12 ms 12 ms 12 ms core4-hu0-16-0-3.faraday.ukcore.bt.net [195.99.127.206]
6 12 ms 12 ms 11 ms peer1-xe-1-0-1.faraday.ukcore.bt.net [213.121.193.197]
7 12 ms 13 ms 12 ms linx-gw1.thn.ncuk.net [195.66.224.240]
8 47 ms 217 ms 129 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
9 12 ms 12 ms 12 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
10 13 ms 13 ms 13 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.

First hop there is 31.55.186.180 so then tried ping -l 1400 -n 100 31.55.186.180

Ping statistics for 31.55.186.180:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 27ms, Average = 25ms

What does that tell us?

Repeat the ping for each node in the chain.

In reply to a post by micksharpe:

Repeat the ping for each node in the chain.

Hop 5: Can't ping it - constant 'request timed out'.
Ping statistics for 195.99.127.206:
Packets: Sent = 84, Received = 0, Lost = 84 (100% loss),

Hop 6: Ping statistics for 213.121.193.197:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 29ms, Average = 26ms

Hop 7: Ping statistics for 195.66.224.240:
Packets: Sent = 100, Received = 99, Lost = 1 (1% loss),
Approximate round trip times in milli-seconds:
Minimum = 26ms, Maximum = 38ms, Average = 26ms

Hop 8: Ping statistics for 80.249.97.17:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 26ms, Maximum = 173ms, Average = 28ms

Hop 9: Ping statistics for 80.249.97.85:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 123ms, Average = 29ms

Hop 10: Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 28ms, Average = 25ms

In reply to a post by dwjoll:

Hop 10: Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 28ms, Average = 25ms

Hmm, so thinkbroadband is now OK, is everywhere OK now, or if not does repeating the hops test show different results?

In reply to a post by legume:

In reply to a post by dwjoll:

Hop 10: Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 28ms, Average = 25ms

Hmm, so thinkbroadband is now OK, is everywhere OK now, or if not does repeating the hops test show different results?

Tracing route to www.thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 12 ms 12 ms 12 ms 31.55.186.180
5 12 ms 11 ms 12 ms core4-hu0-16-0-3.faraday.ukcore.bt.net [195.99.127.206]
6 12 ms 12 ms 12 ms peer1-xe1-0-1.faraday.ukcore.bt.net [213.121.193.197]
7 13 ms 12 ms 16 ms linx-gw1.thn.ncuk.net [195.66.224.240]
8 13 ms 12 ms 12 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
9 13 ms 14 ms 13 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
10 13 ms 12 ms 13 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.

Did tracert again, same IPs.

Now doing this again: ping -l 1400 -n 100 www.thinkbroadband.com

Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 100, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 29ms, Average = 25ms


Very odd.

Speeds have also returned:
http://www.thinkbroadband.com/speedtest/button/14714...

However, I noticed this same thing happening yesterday. All seems well for a short moment until things go downhill again...

When its bad repeat this, as if the ping is fine to that first hop (which will change now and then) then it shows the broadband side is fine, and there may be an issue in the core/backhaul networks.

Not something I'd suggest someone does unless they understand the risks and how to mitigate them by placing a device in the DMZ

In reply to a post by MrSaffron:

When its bad repeat this, as if the ping is fine to that first hop (which will change now and then) then it shows the broadband side is fine, and there may be an issue in the core/backhaul networks.

It just went bad again so tried running the tests.

ping -l 1400 -n 100 www.thinkbroadband.com

Ping statistics for 80.249.99.130:
Packets: Sent = 100, Received = 81, Lost = 19 (19% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 32ms, Average = 26ms

Tracing route to www.thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 12 ms 12 ms 11 ms 31.55.186.180
5 14 ms 12 ms 12 ms core4-hu0-16-0-3.faraday.ukcore.bt.net [195.99.127.206]
6 12 ms 11 ms * peer1-xe1-0-1.faraday.ukcore.bt.net [213.121.193.197]
7 12 ms 13 ms 13 ms linx-gw1.thn.ncuk.net [195.66.224.240]
8 13 ms 12 ms 13 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
9 14 ms 13 ms 12 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
10 * 12 ms * www.thinkbroadband.com [80.249.99.130]
11 12 ms 13 ms 13 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.

Running ping -l 1400 -n 100 tests

Ping to router fine.

Hop 4: This time I noticed packet loss. Same IP as earlier test.

Ping statistics for 31.55.186.180:
Packets: Sent = 100, Received = 70, Lost = 30 (30% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 30ms, Average = 26ms

So this shows there is intermittent packet loss to the first hop. Where would this first hop likely be in the chain, location wise? In the apartment block?

Edited by deleted (Wed 17-Aug-16 22:33:30)

In reply to a post by MrSaffron:

When its bad repeat this, as if the ping is fine to that first hop (which will change now and then) then it shows the broadband side is fine, and there may be an issue in the core/backhaul networks.

Olympics?

In reply to a post by RobertoS:

In reply to a post by MrSaffron:

When its bad repeat this, as if the ping is fine to that first hop (which will change now and then) then it shows the broadband side is fine, and there may be an issue in the core/backhaul networks.

Olympics?

I doubt it.

Anyhow, it's 12:30am now and this is still happening:

Tracing route to www.thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * * * Request timed out.
4 11 ms 12 ms 12 ms 31.55.186.180
5 11 ms 12 ms * core4-hu0-16-0-3.faraday.ukcore.bt.net [195.99.127.206]
6 12 ms 12 ms 11 ms peer1-xe-1-0-1.faraday.ukcore.bt.net [213.121.193.197]
7 12 ms 12 ms 12 ms linx-gw1.thn.ncuk.net [195.66.224.240]
8 12 ms 14 ms 12 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
9 12 ms 13 ms 14 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
10 * 13 ms 16 ms www.thinkbroadband.com [80.249.99.130]

Trace complete.


Ping statistics for 31.55.186.180:
Packets: Sent = 100, Received = 71, Lost = 29 (29% loss),
Approximate round trip times in milli-seconds:
Minimum = 25ms, Maximum = 35ms, Average = 27ms

I don't know if this has any meaning, but the next tracert I did found an extra IP address at hop 3.

Tracing route to www.thinkbroadband.com [80.249.99.130]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms BThomehub.home [192.168.1.254]
2 * * * Request timed out.
3 * 11 ms 11 ms 31.55.186.181
4 13 ms 13 ms 13 ms 31.55.186.180
5 * 13 ms 13 ms 195.99.127.56
6 13 ms 13 ms 13 ms host213-121-193-197.ukcore.bt.net [213.121.193.197]
7 13 ms 13 ms 13 ms linx-gw1.thn.ncuk.net [195.66.224.240]
8 * 12 ms 12 ms te2-1-9.star10g.bdr-rt3.thdo.ncuk.net [80.249.97.17]
9 14 ms 21 ms 15 ms po4-31.core-rs2.thdo.ncuk.net [80.249.97.85]
10 14 ms 14 ms 14 ms www.thinkbroadband.com [80.249.99.130]

Each hop is three ping replies from a single router at that point.

"Request timed out" has two meanings so far as I know. First, the router itself is set not to respond to them, or second the router is very busy and simply stops replying to them at such times. It has things to do it considers more important than replying to pings. That popping up at hop three suggests the second.

Very occasionally I see a mix on a hop of asterisks and times, which comes under the second scenario but it has time to respond to one or two of the three.

If a hop has a high figure compared to what came before but the later hops go back to similar to what came before that just means that router responds slowly for whatever reason. There is no problem. However if the high ping continues into all subsequent hops you gave found a bottle-neck.

Lol at the first hop being in the apartment block.

The first hop is deep in the network, so includes your DSL connection and potentially many miles of fibre core network.

Olympic (i.e. congestion) issues look unlikely as latency is not higher overall, i.e. should rise in latency before dropping packets, look more like an issue somewhere, and figuring this out is down to the BT Core and DSL teams.