Swithced to talktalk. Is this latency typical?

Hi All

I switched to talktalk from virgin yesterday. Please see below my latency graph generated from TBB.

(Huge red bar is downtime while installing, before red bar- virgin with SuperHub 2AC in modem mode, after red bar, TalkTalk)

I am using a Draytek Vigor 130 hooked up to a Linksys WRT1900ACS. My speeds are hovering around 46Mbps/18Mbps Up/Down even though I subscribed for the 80/20 service and my sync shows the max (79999/19999).


https://www.thinkbroadband.com/broadband/monitoring/...

My main concern is the dropped packets and low speeds. Is this because the line needs a few days to stabilise or should I expect this service at all times (and sometimes worse)?

Thanks

Dropped or just delayed?

If you were using the connection during 4pm to midnight then the yellow is most likely just down to your download or upload activity i.e. what you are doing does have an impact on latency

It is when you see the red during idle periods that you have problems.

Low speeds - FTTC should start of at the maximum possible and only slow things down if there is problems, so would need to see speed tests. FTTC does NOT start with slow speeds and then climb up.

I wouldn't be too concerned about your latency spike. In terms of packet loss there is a downward spike at 2am, but it doesn't look like a loss of sync so may just be DLM resetting itself.

Your average ping times are quite high for FTTC (although this may just be dependent on geography), but I'd definitely go into the modem to look at the DSL stats and post them here.

It may well be that you have some interleaving applied, which may partially explain your speed test results.

Please also post your TBB results graph (as MrS has requested) as that will provide a greater insight into what is happening with your line.

Minimum latency looks better on TalkTalk than Virginmedia.
Can you run a speedtest at speedtest.net and let us know the results for a comparison?
Also run a test over here and post the results:
https://www.thinkbroadband.com/speedtest

2AM dropped packets is likely due to DLM, which runs in the evening, or a download of some sorts.

Nothing bad here in my view, looks very good and clean.

Thanks Andrew!

Maybe delayed? I was referring to the little red dots (and sometimes little spikes) between 4PM - midnight (which is while I was mainly using the line)

Regarding yellow spikes, I think you're right, mainly usage impacting latency.

However, we were def sleeping when that 2am big spike happened. I've checked the modem this morning and it said uptime 18hrs. Router said same thing too.

Do you think I should file a low speed complaint? Here is the speedtest I run yesterday:

https://www.thinkbroadband.com/speedtest/15167300028...

What do you mean by the DLM resetting itself? Would that cause a downtime? I wasn't using the connection at that time.

The ping times I believe are higher in Northern Ireland than other areas in the U.K. Here are my modem stats:

---------------------- ATU-R Info (hw: annex A, f/w: annex A/B/C) -----------
Running Mode : 17A State : SHOWTIME
DS Actual Rate : 79995000 bps US Actual Rate : 19999000 bps
DS Attainable Rate : 80373384 bps US Attainable Rate : 25659000 bps
DS Path Mode : Fast US Path Mode : Fast
DS Interleave Depth : 1 US Interleave Depth : 1
NE Current Attenuation : 15 dB Cur SNR Margin : 5 dB
DS actual PSD : 6. 5 dB US actual PSD : 13. 3 dB
NE CRC Count : 1604 FE CRC Count : 39
NE ES Count : 890 FE ES Count : 37
Xdsl Reset Times : 0 Xdsl Link Times : 1
ITU Version[0] : b5004946 ITU Version[1] : 544e0000
VDSL Firmware Version : 05-07-06-0D-01-07 [with Vectoring support]
Power Management Mode : DSL_G997_PMS_L0
Test Mode : DISABLE
-------------------------------- ATU-C Info ---------------------------------
Far Current Attenuation : 17 dB Far SNR Margin : 14 dB
CO ITU Version[0] : b5004244 CO ITU Version[1] : 434da48c
DSLAM CHIPSET VENDOR : < BDCM >


Here is more stats:


Near End Far End Note
Trellis : 1 1
Bitswap : 1 1
ReTxEnable : 0 0
VirtualNoise : 0 0
20BitSupport : 0 0
LatencyPath : 0 0
LOS : 0 0
LOF : 0 0
LPR : 0 0
LOM : 0 0
SosSuccess : 0 0
NCD : 0 0
LCD : 0 0
FECS : 47621 307 (seconds)
ES : 890 37 (seconds)
SES : 0 0 (seconds)
LOSS : 0 0 (seconds)
UAS : 34 12641 (seconds)
HECError : 0 0
CRC : 1604 39
RsCorrection : 0 0
INP : 0 0 (symbols)
InterleaveDelay : 0 0 (1/100 ms)
NFEC : 254 255
RFEC : 16 16
LSYMB : 5374 21451
INTLVBLOCK : 127 255
AELEM : 0 ----



I noticed that G.INP is disabled on the line (ReTxEnable) but I see that my line is on a Fast (not interleaved) path.

The speed test results are in my reply to Andrew. Would appreciate some insight!

Thanks a lot!

Top reason I switched away from Virgin was fluctuating latency. The irony is that very expensive package (VIVID 200 Gamer) is 'tailored' for gamers. It was too late when I realised that the latency problems aren't just coming from the SuperHub 3, but in fact the nature of their network.

Please see my test results in my reply to Andrew. Thanks!

For reference, in addition to TBB's result, here is my result from SpeedTest (over Ethernet as well)

http://www.speedtest.net/result/6995508639

This is even below the downstream hand-back threshold (55mbps). (see here: https://imgur.com/a/9z4Ss)

I believe I have only until next week to cancel. After that I'll be stuck with them for 18 months. Any ideas?

Do a speedtest with the PC directly connected to the Vigor 130 to rule out issues with the Linksys. You may need to setup a PPPoE connection on your PC in order to do this. Also ensure the firmware on the Vigor 130 is up to date.

Can you connect the supplied talktalk router and do speedtests over ethernet. TalkTalk will not acknowledge a fault until this is done anyway.

Can you also post stats from the talktalk router please?

I've just migrated and had a similar blip, which my ISP put down to DLM. I note UKHardy has also suggested it was DLM triggered. The workings of DLM are a bit of a mystery, so I for one are unable to explain it further.

The good news from your stats is that (as you have already spotted) there is no interleaving on your line, although your Errored Seconds count is a bit higher than ideal.

I'd certainly be having words with TT about why your speed test results are as low as they are.

Sorry for the false alarm. I just realised that my SQM (Smart Queue Management) QoS settings on my WRT1900ACS (with LEDE OpenWRT) were still ones leftover from the Virginmedia cable connection. There are significant differences especially when it comes to Link Layer Adaptation settings, which I have now corrected.

The speed looks normal now! (even over WiFi)


With corrected SQM QoS settings: https://www.thinkbroadband.com/speedtest/15168015122...

With SQM QoS Disabled: https://www.thinkbroadband.com/speedtest/15168018476...

Thanks everyone.

In any case, do my line stats still look okay?

BQM is looking better, now to see how it stands up what ever is peak time in your place

Yes! Still monitoring the connection here. Do you (or anyone here) knows by any chance what protocol TalkTalk is using exactly?

It's definitely not PPPoE or PPPoA. In DrayTek configurations, I had to use MPoA/Dynamic IP to get a sync. Any idea what does this mean or how does it work?

Thanks

In reply to a post by nidstigator:

Do you (or anyone here) knows by any chance what protocol TalkTalk is using exactly?

They use DHCP, routers seem to call it Dynamic IP or IPoE.

Hi guys,

I recently joined talktalk myself.
Currently facing extremely poor WhatsApp video calls to and from other talktalk fibre connections.. this goes away if use VPN or my mobile data.
Did some tracerts and pings to whatsapp.com and yes pings over 250ms direct and 80ms via VPN. Traceroute showed -4 routes on VPN and less timing too

Is this a talktalk issue with their weird routing, any ideas how to cure it apart from living VPN always on which I don't really want to do as it affects Netflix and prime video in the house..

kannanni -

Have you tried using different DNS's. ?

i.e. anything other than TalkTalk's own.

In reply to a post by kannanni:

Hi guys,

I recently joined talktalk myself.
Currently facing extremely poor WhatsApp video calls to and from other talktalk fibre connections.. this goes away if use VPN or my mobile data.
Did some tracerts and pings to whatsapp.com and yes pings over 250ms direct and 80ms via VPN. Traceroute showed -4 routes on VPN and less timing too

Is this a talktalk issue with their weird routing, any ideas how to cure it apart from living VPN always on which I don't really want to do as it affects Netflix and prime video in the house..

I am using Cloudflares DNS and am getting the same results.
https://www.cloudflare.com/learning/dns/what-is-1.1....


Ping results - https://i.postimg.cc/9M6PVw8f/ping.jpg

Added this in case its of use - https://i.postimg.cc/PJfS0wP8/trace.jpg

Edit to add second link.

Edited by deleted (Tue 07-Jan-20 15:58:57)

I use cloudflare and Google DNS too.
This is not related to dns. It's just poor talktalk routing/peering I'm afraid..talktalk support thinks this is normal and cannot fix it. Virgin,bt,hyperoptic I tested today all return 70-80ms to WhatsApp servers.
I resolved using duo for now for video calls and it's better..

Well I joined TalkTalk FTTC 4 days ago and I get very low pings.

C:\Windows\system32>ping whatsapp.com

Pinging whatsapp.com [169.55.60.148] with 32 bytes of data:
Reply from 169.55.60.148: bytes=32 time=77ms TTL=51
Reply from 169.55.60.148: bytes=32 time=77ms TTL=51
Reply from 169.55.60.148: bytes=32 time=76ms TTL=51
Reply from 169.55.60.148: bytes=32 time=77ms TTL=51

Ping statistics for 169.55.60.148:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 76ms, Maximum = 77ms, Average = 76ms

C:\Windows\system32>ping lichess.org

Pinging lichess.org [37.187.205.99] with 32 bytes of data:
Reply from 37.187.205.99: bytes=32 time=6ms TTL=54
Reply from 37.187.205.99: bytes=32 time=6ms TTL=54
Reply from 37.187.205.99: bytes=32 time=6ms TTL=54
Reply from 37.187.205.99: bytes=32 time=7ms TTL=54

Ping statistics for 37.187.205.99:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 7ms, Average = 6ms

C:\Windows\system32>ping bbc.com

Pinging bbc.com [151.101.64.81] with 32 bytes of data:
Reply from 151.101.64.81: bytes=32 time=4ms TTL=59
Reply from 151.101.64.81: bytes=32 time=3ms TTL=59
Reply from 151.101.64.81: bytes=32 time=3ms TTL=59
Reply from 151.101.64.81: bytes=32 time=4ms TTL=59

Ping statistics for 151.101.64.81:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 3ms, Maximum = 4ms, Average = 3ms

Generally speaking 3-6ms ping across all of these tests. In speed tests I also get 4-5ms ping consistently here in London, Bishopsgate.

I think latency may be area specific. This is the first time I've had pings this low. With ADSL Plusnet was 20ms and with Sky it was 12ms. Lichess.org I was getting 24ms ping with Sky/Plusnet ADSL. But with TalkTalk FTTC so far it has been 6ms ping which is a massive difference.

your WhatsApp ping is 10x higher though,.....

With all due respect, it doesn't matter if only WhatsApp ping is so high because I don't care about it.

What's important is that the overall pings are low throughout all websites. And in my case that is all that matters.

I just tested Viber now C:\Windows\system32>ping viber.com

Pinging viber.com [2.22.132.47] with 32 bytes of data:
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51

Ping statistics for 2.22.132.47:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 6ms, Average = 6ms

Viber is an alternative program that can be used instead of WhatsApp and in this case I am pinging at 6ms response time, which is great.

I am also pinging Facebook at 4-6ms which is also something I use for voice chat.

I would only complain if I had 77ms in majority of the websites. But if the majority pings at 3-6ms then I have nothing to complain about!

This thread is somewhat confusing because it creates the impression that the latencies are high throughout the entire TalkTalk ping tests. And this is simply not true, at least not in my case.

The OP is also complaining about packet loss. But as you can see from my tests there are no packets lost in any of my ping tests.

In reply to a post by Myth:

your WhatsApp ping is 10x higher though,.....

Not that surprising given the server the request went to is in the United States. Light can only travel so fast.

In reply to a post by CarlTSpeak:

In reply to a post by Myth:

your WhatsApp ping is 10x higher though,.....

Not that surprising given the server the request went to is in the United States. Light can only travel so fast.

Well I am not sure if this is necessarily true. Because as we can see https://ipinfo.io/2.22.132.47

2.22.132.47 Cambridge, Massachusetts, United States

C:\Windows\system32>ping viber.com

Pinging viber.com [2.22.132.47] with 32 bytes of data:
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51
Reply from 2.22.132.47: bytes=32 time=6ms TTL=51

Ping statistics for 2.22.132.47:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 6ms, Average = 6ms

Viber.com pings are displaying for me 6ms from a US server, how do you explain that?

That IP address is on the Akamai Content Delivery Network so could be pretty much anywhere. The geolocation is meaningless, that's just where Akamai are registered as a business.

Hope that helps.

In reply to a post by CarlTSpeak:

That IP address is on the Akamai Content Delivery Network so could be pretty much anywhere. The geolocation is meaningless, that's just where Akamai are registered as a business.

Hope that helps.

Here ya go:

Non-authoritative answer:
Name: viber.com
Address: 2.22.132.47

Non-authoritative answer:
47.132.22.2.in-addr.arpa name = a2-22-132-47.deploy.static.akamaitechnologies.com.

traceroute to 2.22.132.47 (2.22.132.47), 64 hops max, 52 byte packets
1 192.168.0.1 (192.168.0.1) 3.692 ms 2.068 ms 1.060 ms
2 10.93.56.1 (10.93.56.1) 10.937 ms 9.051 ms 10.852 ms
3 leed-core-2b-xe-812-0.network.virginmedia.net (80.0.241.177) 10.236 ms 11.965 ms 13.534 ms
4 * * *
5 * * *
6 86.85-254-62.static.virginmediabusiness.co.uk (62.254.85.86) 27.026 ms 24.256 ms 33.220 ms
7 * * *
8 uk-lon03a-ri1-ae-2-0.aorta.net (84.116.135.46) 21.572 ms 21.992 ms 21.313 ms
9 ae11-209.rt.tc2.lon.uk.retn.net (87.245.245.22) 21.352 ms 20.985 ms 42.402 ms
10 gw-akamai.retn.net (87.245.245.29) 23.362 ms 21.736 ms 23.188 ms
11 * * *

For those of us in the UK t'is in Londonium

Viber.com pings are displaying for me 6ms from a US server, how do you explain that?

He did explain it below.

I'd much rather hear your explanation of how your connection can break the laws of physics.

Never seen 6ms from Scotland to London, nevermind London to Massachusetts.

Well here's a bit of potentially useless knowledge.

Back in the day when I was involved in the sale of fibre inter-connectivity for remote data centres we used a rule of thumb of 1 ms for every 150 miles.

A lot of things have changes since then, but I suspect not the speed of light.

Edited by iefbr14 (Fri 14-Feb-20 12:28:01)

Fibre with better refractive index is a thing but I'm not sure how many production networks, if any, it is in.

High frequency traders continue to use microwave links in preference to fibre for good reasons. Still getting about 2/3rds c from fibre while microwave gets you really close to it.

In reply to a post by BLaZiNgSPEED:

With all due respect, it doesn't matter if only WhatsApp ping is so high because I don't care about it.

....

......
The OP is also complaining about packet loss. But as you can see from my tests there are no packets lost in any of my ping tests.

I see your issue, we're currently not replying to the OP, but to someone who resurrected this thread this year, who specifically is investigating an issue with WhatsApp

Edited by deleted (Sun 16-Feb-20 10:10:02)

are you still getting ~80ms with the 169.55.60.148 IP?

I'm getting ~300ms with both the .148 and the original .170

In reply to a post by Myth:

are you still getting ~80ms with the 169.55.60.148 IP?

I'm getting ~300ms with both the .148 and the original .170

You are right, it has gone bonkers again!

Pinging whatsapp.com [169.55.60.170] with 32 bytes of data:
Reply from 169.55.60.170: bytes=32 time=244ms TTL=52
Reply from 169.55.60.170: bytes=32 time=244ms TTL=52
Reply from 169.55.60.170: bytes=32 time=244ms TTL=52
Reply from 169.55.60.170: bytes=32 time=244ms TTL=52

Ping statistics for 169.55.60.170:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 244ms, Maximum = 244ms, Average = 244ms

C:\Windows\system32>ping 169.55.60.148

Pinging 169.55.60.148 with 32 bytes of data:
Reply from 169.55.60.148: bytes=32 time=225ms TTL=50
Reply from 169.55.60.148: bytes=32 time=225ms TTL=50
Reply from 169.55.60.148: bytes=32 time=225ms TTL=50
Reply from 169.55.60.148: bytes=32 time=225ms TTL=50

Ping statistics for 169.55.60.148:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 225ms, Maximum = 225ms, Average = 225ms

But this is about the only website that is causing this problem. It was 78ms before but has now jumped to 224ms. Fortunately I don't care about WhatsApp as it doesn't affect me. I care only about a few sites like lichess.org Facebook, Youtube, Steam game server which are all under 6ms. Including Viber.

Ok, something very interesting. Whatsapp.com seemed to have moved servers and is now reporting lower pings!

C:\Windows\system32>ping whatsapp.com

Pinging whatsapp.com [157.240.1.53] with 32 bytes of data:
Reply from 157.240.1.53: bytes=32 time=4ms TTL=55
Reply from 157.240.1.53: bytes=32 time=4ms TTL=55
Reply from 157.240.1.53: bytes=32 time=5ms TTL=55
Reply from 157.240.1.53: bytes=32 time=4ms TTL=55

Ping statistics for 157.240.1.53:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 4ms, Maximum = 5ms, Average = 4ms

Since yesterday, just out of curiosity I have decided to ping whatsapp, not that I really care about it at all. And it is now pinging at 4-5ms under the new IP address 157.240.1.53. https://whatismyipaddress.com/ip/157.240.1.53 https://ipinfo.io/157.240.1.53

The most bizarre thing is that this new address is still located in the United States. Los Angeles, California.

The old IP address still pings at Ping statistics for 169.55.60.148:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 232ms, Maximum = 233ms, Average = 232ms

https://whatismyip.live/ip/169.55.60.148
So the old IP address for Whatsapp.com that was located in Dallas is pinging at 200+ms. And the same goes with 169.55.60.170

But the new IP address that is under California is pinging at 4-5ms? How comes? I really do not know.

And yes, I am also a TalkTalk FTTC customer.

It is called content delivery networks that mirror popular sites around the globe so you access a version that is closest to you.

You would be changing the laws of physics to get that ping time to the states, it has to be a local site as MrSaffron has explained.

The IP address might show was being in California but with a 4ms response that's certainly not where you're pinging.

Running a tracert to WhatsApp.com ends up with me at whatsapp-cdn-shv-01-ams4.fbcdn.net.

It's there in the hostname, it's a WhatsApp/Facebook CDN and it's based in the UK.

Ok, many thanks, now this makes perfect sense! I too was finding it strange to see it pinging at 4ms initially. I'm aware a ping cannot be that low coming from the US.

Could it be that WhatsApp.com has finally moved to UK servers?

Last 2 months since being a TalkTalk customer, WhatsApp.com has always been mostly 200ms ping with the exception on one occasion when it was 76ms.

Some of the users in this thread were complaining that their pings with WhatsApp.com were too high with TalkTalk something I too have noticed.

Hence why I decided to inform of this new change that I am now seeing with the pings being lower this time around 4-5ms for the first time.

I'm guessing the main concern of having those high pings were that they were contributing to latency delays in video and voice calls? Now with the lower pings the WhatsApp performance should have modest improvements.

You are of course assuming that audio and video is using the whatsapp.com IP address.

An app or webpage can pull/send video/audio to many different places beyond the web page you see in the URL of the page.