gtoderici's profile

Frequent Visitor

 • 

9 Messages

Thursday, March 4th, 2021 5:00 PM

Closed

horrendous latency issues (peak of 8 seconds to google.com) - need help

I have a 200mbps plan, and I have been experiencing some terrible latency when using *WIRED* connections. I ran mtr over a period of several hours pinging google.com, and observed that:

 

1. The latency to the router is around 2ms

2. The latency to the next hop after the modem is ... 4-8 SECONDS... I reset both modem/router and got the same issue. This happened today again (it has happened in multiple days). The standard deviation for latency during the day is around 1 second these days, which is unacceptable. 

3. The packet drop rate is around 0.1% across the entire day (I guess this has happened multiple times during my meetings).

4. Throughput is not alwyas an issue but sometimes it is. I saw it go from 200mbps to .... 500kbps sometimes. This is not repeatable, though, and it's not always (though it does coincide sometimes) at the same time as the latency spikes.

 

For reference, I have a Arris SB8200 modem, and tried TWO routers which didn't seem to matter. One was Asus rt-ax88u, and the other was a Netgear Nighthawk X4 R7500.

 

I have another house nearby to where this is happening (~1 mile away), and I am NOT seeing these issue there with the same equipment.

 

It's driving me nuts because when I am in a video call, SOMETIMES I will get hit by the 8 second latency, and at that time it feels like I am talking to someone from the moon due to the insane latency.

 

What is mega-frustrating is that I got to speak to some Xfinity specialist on the phone previously about this and the summary of the conversation was that I have a problem with "milliseconds". As an outcome of that a tech came by and mentioned that maybe the issue is happening is with the node to which the modem is connected to and that he would open a ticket, but I haven't heard back.

 

I am getting desperate because due to COVID-19 I am forced to work from home, and I am in meetings >75% of the day, and currently with the connection I have it's getting really really uncomfortable to meet anyone (dropped audio/super-huge lag as I described above).

Frequent Visitor

 • 

9 Messages

4 years ago

This is what I see when running mtr google.com right now (off-peak, ~6pm my time zone). I ran below for about half an hour and saw spikes of over 200ms as highlighted.

Please notice the packet loss which somehow is around 0.3% to google.com. THIS IS ON A WIRED CONNECTION.


Host Loss% Snt Last Avg Best Wrst StDev
1. 192.168.1.1 0.1% 2934 0.6 0.5 0.4 2.9 0.2
2. 96.120.88.113 0.2% 2934 27.5 11.8 4.4 249.9 10.3
3. 96.110.177.13 0.1% 2934 10.8 12.2 4.8 296.1 12.6
4. 162.151.78.129 0.2% 2934 10.1 12.4 7.1 295.9 13.3
5. be-232-rar01.santaclara.ca.sfba.comcast.net 0.1% 2934 26.4 14.7 7.7 289.3 13.9
6. be-299-ar01.santaclara.ca.sfba.comcast.net 0.2% 2934 10.8 15.3 7.7 312.0 22.6
7. 96.112.146.18 0.2% 2933 15.4 11.9 7.6 286.8 9.7
8. 74.125.37.224 0.3% 2933 15.3 13.3 9.2 209.6 9.8
9. 72.14.235.3 0.3% 2933 12.6 13.4 9.2 347.2 10.7
10. sfo03s08-in-f14.1e100.net 0.3% 2933 12.5 13.2 9.2 383.7 10.7

Frequent Visitor

 • 

9 Messages

4 years ago

This is what the modem is reporting currently (no major issue RIGHT NOW).
 

The status listed show the connection state of the cable modem. They are used by your service provider to evaluate the operation of the cable modem.

Startup Procedure
ProcedureStatusComment
Acquire Downstream Channel507000000 HzLocked
Connectivity StateOKOperational
Boot StateOKOperational
Configuration FileOK 
SecurityEnabledBPI+
DOCSIS Network Access EnabledAllowed 

 
Downstream Bonded Channels
Channel IDLock StatusModulationFrequencyPowerSNR/MERCorrectedUncorrectables
20LockedQAM256507000000 Hz2.8 dBmV39.6 dB00
17LockedQAM256483000000 Hz3.0 dBmV40.0 dB00
18LockedQAM256489000000 Hz3.1 dBmV39.9 dB00
19LockedQAM256495000000 Hz2.9 dBmV39.8 dB00
21LockedQAM256513000000 Hz2.6 dBmV39.7 dB00
22LockedQAM256519000000 Hz2.3 dBmV39.6 dB00
23LockedQAM256525000000 Hz2.2 dBmV39.4 dB00
24LockedQAM256531000000 Hz1.9 dBmV39.0 dB00
25LockedQAM256543000000 Hz2.2 dBmV39.5 dB00
26LockedQAM256549000000 Hz2.4 dBmV39.5 dB00
27LockedQAM256555000000 Hz2.6 dBmV39.6 dB00
28LockedQAM256561000000 Hz2.5 dBmV39.4 dB00
29LockedQAM256567000000 Hz2.3 dBmV39.5 dB00
30LockedQAM256573000000 Hz1.8 dBmV39.2 dB00
31LockedQAM256579000000 Hz1.6 dBmV39.4 dB00
32LockedQAM256585000000 Hz1.6 dBmV39.0 dB00
33LockedQAM256591000000 Hz1.4 dBmV39.1 dB00
34LockedQAM256597000000 Hz1.2 dBmV39.1 dB00
35LockedQAM256603000000 Hz0.9 dBmV38.5 dB00
36LockedQAM256609000000 Hz0.8 dBmV38.7 dB00
37LockedQAM256615000000 Hz0.6 dBmV38.7 dB00
38LockedQAM256621000000 Hz0.6 dBmV38.9 dB00
39LockedQAM256627000000 Hz0.6 dBmV38.6 dB00
40LockedQAM256633000000 Hz0.6 dBmV38.5 dB00
41LockedQAM256639000000 Hz0.4 dBmV38.0 dB00
42LockedQAM256645000000 Hz0.6 dBmV38.6 dB00
43LockedQAM256651000000 Hz0.3 dBmV38.4 dB00
44LockedQAM256657000000 Hz0.3 dBmV38.2 dB00
45LockedQAM256663000000 Hz0.2 dBmV38.2 dB00
46LockedQAM256669000000 Hz-0.4 dBmV38.1 dB00
47LockedQAM256675000000 Hz-1.0 dBmV37.4 dB00
48LockedOther688000000 Hz-1.0 dBmV34.5 dB8741838660



 
Upstream Bonded Channels
ChannelChannel IDLock StatusUS Channel TypeFrequencyWidthPower
11LockedSC-QAM Upstream16400000 Hz6400000 Hz41.0 dBmV
22LockedSC-QAM Upstream22800000 Hz6400000 Hz39.0 dBmV
33LockedSC-QAM Upstream29200000 Hz6400000 Hz39.0 dBmV
44LockedSC-QAM Upstream35600000 Hz6400000 Hz38.0 dBmV
55LockedSC-QAM Upstream39600000 Hz1600000 Hz35.0 dBmV



 

Current System Time: Thu Mar 4 17:31:32 2021

Frequent Visitor

 • 

9 Messages

4 years ago

As I pointed out above, 370ms (0.37s) is when I captured the log. I should have taken a screenshot with the 8 second one. There's no confusion on units.

Problem Solver

 • 

948 Messages

4 years ago

1000 ms = 1 second so a spike  of 370 ms is only .37 of a second, out of the 3000 packets you sent the average was .013 of a second, call me crazy but i don't see horrendous latency, I believe the tech told you the problem was "milliseconds" because i believe your confusing milliseconds with seconds which are two different measurements of time

Problem Solver

 • 

948 Messages

4 years ago

@EG  Thank you, you always have the answers, 

Frequent Visitor

 • 

9 Messages

4 years ago

Digging through the modem logs, I also noticed these messages around the time of super-high latency:

 

Event Log

The table below contains the log of events that the SB8200 has detected. This log can be important to the service provider to help diagnose and correct problems, if any should occur.

Date Time Event ID Event Level Description
03/04/2021 15:10670616006"DS profile assignment change. DS Chan ID: 32; Previous Profile: ; New Profile: 1 2 3.;CM-MAC=6c:63:9c:b3:72:25;CMTS-MAC=00:01:5c:63:fc:57;CM-QOS=1.1;CM-VER=3.1;"
03/04/2021 15:10680011023"Config File Rejected due to Invalid or Unexpected TLV 11 encoding;CM-MAC=6c:63:9c:b3:72:25;CMTS-MAC=00:01:5c:63:fc:57;CM-QOS=1.1;CM-VER=3.1;"
03/04/2021 15:10730402003"TLV-11 - Illegal Set operation failed;CM-MAC=6c:63:9c:b3:72:25;CMTS-MAC=00:01:5c:63:fc:57;CM-QOS=1.1;CM-VER=3.1;"

Problem Solver

 • 

948 Messages

4 years ago

i know nothing about the logs, but when i query google.com i get differnt ips, maybe some load balencing thing i dont know, here it 2 mtr results. this is the google.com you pinged and this is another google.com i ping, very differnt results, also i am curios how you recorded the 8 minute lag, or what told you it was that long

2 Attachments

Frequent Visitor

 • 

9 Messages

4 years ago

I'd like to emphasize that the latency I was experiencing is 8 seconds (not minutes). 

 

I also ran mtr from a two different ISPs today and I noticed that the worst of the words in terms of ping to google.com was never above 200-300ms (depending on which ISP I used, but sadly none of which are available at my home address). 

 

However, I am also not experiencing the horrendous latency today on xfinity, so maybe the problem was transient? This has happened a few times over the past 3 months (I can count about 5-6 times).

 

The problem seems to be hyper-local here where I live because at another property where I also have xfinity and there I don't get such incredible latency spikes (it's very close to the current address ~0.5 miles away). One more time, I am OK with peaks of 300ms, but not 8000ms.

Expert

 • 

110.1K Messages

4 years ago

@jlavaseur 

 

The reason for the difference. The first one is the west coast. Typical destination server RTT considering the cross-country distance from you.

 

The second one is east coast. Typical destination server RTT considering it is is local to you. Apples / oranges.

Expert

 • 

110.1K Messages

4 years ago

@gtoderici Sorry that none of this directly addresses your issue......

Problem Solver

 • 

948 Messages

4 years ago

Sorry I mistyped, it's funny in the late 90s, we would run web servers, dns servers, email servers all on one computer from dialup! It worked just fine, 56k connection, anyway my sister had a older computer that the video would lag during zoom calls, she thought it was the internet connection, well I opened the task manager, set the priority level to a very high one, that allocated more cpu power and memory to her zoom program, fixed her issues 

Expert

 • 

110.1K Messages

4 years ago

Perhaps there is an intermittent capacity / traffic congestion problem on your local cable segment / node. If so, there is not much that you can do about it except to complain persistently ! Good luck !

Frequent Visitor

 • 

9 Messages

4 years ago

Here's the mtr output from this morning, captured over a period of 75 minutes... the highlights being >5 seconds ping  in the worst case, >260ms STANDARD DEVIATION of the ping to all nodes besides my router...

 

 

 

1 Attachment

Problem Solver

 • 

948 Messages

4 years ago

Ok, they look bad, the hops are probably routers with firewalls, so ping flooding them with 4700 ping requests over 75 minutes, you really don't know if it's the network or you got caught up in a firewall 

Expert

 • 

110.1K Messages

4 years ago

The high latency starts between your router and the first hop which is the CMTS at your local headend facility / the Comcast system.

 

FWIW, I'll state this once again;

 

"Perhaps there is an intermittent capacity / traffic congestion problem on your local cable segment / node. If so, there is not much that you can do about it except to complain persistently ! Good luck ! "

forum icon

New to the Community?

Start Here