Community Forum

Latency Variation/Packet Loss, Possible Bufferbloat?

Highlighted
Regular Visitor

Latency Variation/Packet Loss, Possible Bufferbloat?

Hi all, hoping someone could help here, I'm at my wits end trying to troubleshoot this and having a hard time getting in touch with support. I've started here, with no luck (https://forums.xfinity.com/t5/Your-Home-Network/Internet-Troubleshooting-Tips/m-p/3310447#M316652).

 

I'm on a wired connection. Really the only issue I'm having is playing online games ,streaming and normal day to day is just fine. I'm sure it happens throughout the day, but it's obviosuly extremely noticeable when playing games. I'll randomly get massive latency variation or dropped packets all togther. It will last for about 30 or seconds, and it will happen several times in a 10 minute time span. It also seems to be getting worse over the last few weeks. It seems to be related to the uplink., I've noticed when running a speed test and doing a constant ping to google.com, during the download phase I'll see one or two timeouts but the response time will sit around the normal 30-60ms. As soon as the upload phase starts, response times hold around 150-250ms with plenty of timeouts in between. DSLReports also typically gives me a C or D rating with Bufferbloat when running those tests.

 

I've tried several things, I upgraded to an SB8200 modem, I tried a few different routers (Netgear R7400 and AC1200 with both Netgear firmware and OpenWrt so I could try out the SQM plugin). I've updated splitters, coax throughout the house, just about everything.

 

Since I know modem signal stats will be asked for, here they are...

 
Downstream Bonded Channels
Channel ID Lock Status Modulation Frequency Power SNR/MER Corrected Uncorrectables
21 Locked QAM256 577750000 Hz 4.2 dBmV 40.7 dB 0 0
1 Locked QAM256 457750000 Hz 3.7 dBmV 40.3 dB 0 0
2 Locked QAM256 463750000 Hz 4.1 dBmV 40.4 dB 0 0
3 Locked QAM256 469750000 Hz 4.4 dBmV 40.5 dB 0 0
4 Locked QAM256 475750000 Hz 4.6 dBmV 40.4 dB 0 0
5 Locked QAM256 481750000 Hz 4.5 dBmV 40.5 dB 0 0
6 Locked QAM256 487750000 Hz 4.5 dBmV 40.4 dB 0 0
7 Locked QAM256 493750000 Hz 4.1 dBmV 40.2 dB 0 0
8 Locked QAM256 499750000 Hz 3.8 dBmV 40.1 dB 0 0
9 Locked QAM256 505750000 Hz 3.7 dBmV 40.2 dB 0 0
10 Locked QAM256 511750000 Hz 3.8 dBmV 40.2 dB 0 0
11 Locked QAM256 517750000 Hz 4.1 dBmV 40.5 dB 0 0
12 Locked QAM256 523750000 Hz 4.4 dBmV 40.6 dB 0 0
13 Locked QAM256 529750000 Hz 4.6 dBmV 40.6 dB 0 0
14 Locked QAM256 535750000 Hz 4.9 dBmV 40.6 dB 0 0
15 Locked QAM256 541750000 Hz 5.1 dBmV 40.6 dB 0 0
16 Locked QAM256 547750000 Hz 5.1 dBmV 40.8 dB 0 0
17 Locked QAM256 553750000 Hz 4.9 dBmV 40.6 dB 0 0
18 Locked QAM256 559750000 Hz 4.6 dBmV 40.5 dB 0 0
19 Locked QAM256 565750000 Hz 4.3 dBmV 40.6 dB 0 0
20 Locked QAM256 571750000 Hz 4.1 dBmV 40.5 dB 0 0
22 Locked QAM256 583750000 Hz 4.3 dBmV 40.5 dB 0 0
23 Locked QAM256 589750000 Hz 4.6 dBmV 40.5 dB 0 0
24 Locked QAM256 595750000 Hz 4.8 dBmV 40.5 dB 0 0
25 Locked QAM256 601750000 Hz 4.9 dBmV 40.5 dB 0 0
26 Locked QAM256 607750000 Hz 4.7 dBmV 40.5 dB 0 0
27 Locked QAM256 613750000 Hz 4.3 dBmV 40.3 dB 0 0
28 Locked QAM256 619750000 Hz 4.0 dBmV 40.2 dB 0 0
29 Locked QAM256 625750000 Hz 3.6 dBmV 40.1 dB 0 0
30 Locked QAM256 631750000 Hz 3.2 dBmV 39.6 dB 0 0
31 Locked QAM256 637750000 Hz 3.5 dBmV 39.9 dB 0 0
32 Locked QAM256 643750000 Hz 3.7 dBmV 39.9 dB 0 0
159 Locked Other 701000000 Hz 5.2 dBmV 39.2 dB 306783451 0
Upstream Bonded Channels
Channel Channel ID Lock Status US Channel Type Frequency Width Power
1 3 Locked SC-QAM Upstream 23700000 Hz 6400000 Hz 42.0 dBmV
2 1 Locked SC-QAM Upstream 36500000 Hz 6400000 Hz 42.0 dBmV
3 2 Locked SC-QAM Upstream 30100000 Hz 6400000 Hz 43.0 dBmV

 

As for the logs, I do see a lot of "Dynamic Range Window violations", other than that it's just T3 timeouts here and there...
 
04/03/2020 17:43 2436694061 5 "Dynamic Range Window violation"
04/03/2020 17:43 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:39 2436694061 5 "Dynamic Range Window violation"
04/03/2020 17:39 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:21 2436694061 5 "Dynamic Range Window violation"
04/03/2020 17:21 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:17 2436694061 5 "Dynamic Range Window violation"
04/03/2020 17:17 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:11 2436694061 5 "Dynamic Range Window violation"
04/03/2020 17:11 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:07 82000200 3 "No Ranging Response received - T3 time-out;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 17:07 84000500 3 "SYNC Timing Synchronization failure - Loss of Sync;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 16:55 2436694061 5 "Dynamic Range Window violation"
04/03/2020 16:55 82001100 5 "RNG-RSP CCAP Commanded Power Exceeds Value Corresponding to the Top of the DRW;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"
04/03/2020 16:55 82000200 3 "No Ranging Response received - T3 time-out;CM-MAC=14:c0:3e:0b:a0:ed;CMTS-MAC=00:5f:86:92:22:b2;CM-QOS=1.1;CM-VER=3.1;"

 

Any help is appreciated!

 

Highlighted
Regular Contributor

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

The issue is consistent with another device on your home network doing a massive upload for minutes at a time and causing your game performance to suffer from head-of-line-blocking.      This situation can be mitigated on many routers via specific awesome QOS functionalities involving prioritzaion for well-known ports and/or deep-packet-inspection, video vs. voice vs. sftp for example.  An older-school more-brutish-force way to mitigate head-of-line-blocking is to lower the layer 3 MTU much smaller than the layer 2 MTU.   Various routers provide support for layer 3 MTU adjustment.    Lowering MTU will reduce your throughput rate for big uploads and downloasd but for gaming it may be a good tradeoff.    Gamers will know the good MTU settings , if indeed anyone messes with MTU any more.   

 

Highlighted
Regular Visitor

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

Thanks for the suggestion bald. I have very few devices on my network, and since this also occurs when the router is out of the picture and I'm directly connected to the modem, I'm not quite sure it's some other device uploading something. I'll mess around with MTU a bit and see if that helps.
Highlighted
Regular Contributor

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

You are most welcome.   If the issue happens when you are hardwire to the modem then my ideas above are definitely not appropriate, sorry about that and that I haven't helped you diagnose the issue yet.   An easy test:  are you sure the hardwire is good, it's so inexpensive, swap it with another if possible? 

 

It is usually best to first determine which layer is dropping the packet, top-down.   Can you investigate/find/run the "traceroute" program/app on your device or any device that reproduces the problem?  If so, maybe consider to compare/show-us  the outputs from when problem is happening vs not happening - this can sometimes show the external network-hop that is troublesome.      You can post traceroute outputs and I will help you interpret the weird parts or there are good tutorials online too.   

 

For other testing inside your house, a possibly-good goal for you would be to try to disprove the possibility that it is an App issue or a computer issue: reproduce the problem on hard-wire using a different app/game.   Reproduce the problem on hard-wire using a different game-console/computer/device, or a different OS on the same device.   Definitely try two different hard-wires.

 

 Those aberrant ping times are horrendous but it is interesting that the pings do succeed with such slow turnaround.   That symptom indicates a layer 3 problem too - traceroute can help to show it.   if you have windows you can probably find it under powershell/CMD.EXE/COMMAND.COM.   

Highlighted
Regular Visitor

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

Thanks again bald! I have indeed swapped out/bought new cables, both coax and cat.

 

I've went ahead and tried a few more things. I've hardwired two different PCs, both directly to the modem. My ping tests came up the same on both, huge variation and packet drops. Here's the results of a couple trace routes done in succession while hardwired to the modem...

 

Tracing route to google.com [2607:f8b0:4006:810::200e]
over a maximum of 30 hops:

1 * * * Request timed out.
2 28 ms 15 ms 9 ms 2001:558:202:1160::1
4 53 ms 24 ms * 2001:558:200:380::1
5 * 18 ms * be-7015-cr02.newyork.ny.ibone.comcast.net [2001:558:0:f6b6::1]
6 41 ms * * be-10390-pe03.111eighthave.ny.ibone.comcast.net [2001:558:0:f8d6::2]
7 18 ms 38 ms 17 ms 2001:559::b52
8 18 ms 16 ms 17 ms 2607:f8b0:80d9::1
9 20 ms 27 ms 41 ms 2001:4860:0:1::2118
10 18 ms 60 ms 17 ms 2001:4860:0:1::2105
11 43 ms 17 ms 34 ms lga34s14-in-x0e.1e100.net [2607:f8b0:4006:810::200e]

--------------------------------

Tracing route to google.com [2607:f8b0:4006:810::200e]
over a maximum of 30 hops:

1 * * * Request timed out.
2 26 ms 95 ms 12 ms 2001:558:202:1160::1
3 8 ms 11 ms 9 ms 2001:558:200:806::1
4 * * * Request timed out.
5 * * * Request timed out.
6 * * * Request timed out.
7 17 ms 18 ms 42 ms 2001:559::b52
8 17 ms 18 ms 34 ms 2607:f8b0:80d9::1
9 30 ms 17 ms 27 ms 2001:4860:0:1::2118
10 60 ms 30 ms 17 ms 2001:4860:0:1::2105
11 18 ms 16 ms 18 ms lga34s14-in-x0e.1e100.net [2607:f8b0:4006:810::200e]

--------------------------------

Tracing route to google.com [2607:f8b0:4006:810::200e]
over a maximum of 30 hops:

1 * * * Request timed out.
2 8 ms 7 ms 9 ms 2001:558:202:1160::1
3 28 ms 35 ms 15 ms 2001:558:200:806::1
4 * * * Request timed out.
5 * * * Request timed out.
6 * * * Request timed out.
7 17 ms 33 ms 25 ms 2001:559::b52
8 19 ms 20 ms 18 ms 2607:f8b0:80d9::1
9 17 ms * 26 ms 2001:4860:0:1::2118
10 34 ms 17 ms 16 ms 2001:4860:0:1::2105
11 15 ms 18 ms 24 ms lga34s14-in-x0e.1e100.net [2607:f8b0:4006:810::200e]

---------------------------------

Tracing route to google.com [2607:f8b0:4006:802::200e]
over a maximum of 30 hops:

1 * * * Request timed out.
2 6 ms 9 ms 9 ms 2001:558:202:1160::1
3 8 ms 7 ms 10 ms 2001:558:200:806::1
4 15 ms * * 2001:558:200:380::1
5 * * 29 ms be-7015-cr02.newyork.ny.ibone.comcast.net [2001:558:0:f6b6::1]
6 * * * Request timed out.
7 46 ms 22 ms 42 ms 2001:559:0:19::a
8 38 ms 19 ms 16 ms 2001:4860:0:1128::1
9 54 ms 110 ms 154 ms 2001:4860:0:1::2113
10 18 ms 41 ms 33 ms lga34s12-in-x0e.1e100.net [2607:f8b0:4006:802::200e]

----------------------------------

Tracing route to google.com [2607:f8b0:4006:802::200e]
over a maximum of 30 hops:

1 * * * Request timed out.
2 8 ms 8 ms 10 ms 2001:558:202:1160::1
3 12 ms 10 ms 9 ms 2001:558:200:806::1
4 * 53 ms * 2001:558:200:380::1
5 * 39 ms * be-7015-cr02.newyork.ny.ibone.comcast.net [2001:558:0:f6b6::1]
6 * * * Request timed out.
7 76 ms 39 ms 92 ms 2001:559:0:19::a
8 67 ms 55 ms 59 ms 2001:4860:0:1128::1
9 34 ms 22 ms * 2001:4860:0:1::2113
10 31 ms 36 ms 51 ms lga34s12-in-x0e.1e100.net [2607:f8b0:4006:802::200e]

----------------------------------

And just to have another one that isn't google.com nor IPv6...

 

Tracing route to xfinity.com [68.87.41.40]
over a maximum of 30 hops:

1 30 ms 17 ms 45 ms 96.120.64.13
2 26 ms 89 ms 25 ms 96.108.152.153
3 100 ms 25 ms 210 ms 96.108.157.1
4 * 163 ms 107 ms 162.151.191.141
5 101 ms 61 ms 63 ms be-7015-cr02.newyork.ny.ibone.comcast.net [68.86.90.217]
6 136 ms 22 ms 26 ms be-10203-cr01.newark.nj.ibone.comcast.net [68.86.85.185]
7 48 ms 281 ms 36 ms be-7922-ar02-d.westchester.pa.bo.comcast.net [68.86.91.178]
8 45 ms 121 ms 68 ms be-100-ur22-d.westchester.pa.bo.comcast.net [68.85.137.221]

 

Highlighted
Regular Visitor

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

And just as a comparision for when things are going well...

 

Tracing route to google.com [2607:f8b0:4006:804::200e]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 2601:183:c900:23a3:bea5:11ff:fe00:3e52
2 * * * Request timed out.
3 7 ms 8 ms 9 ms 2001:558:202:115f::1
4 * * * Request timed out.
5 * * * Request timed out.
6 * * * Request timed out.
7 19 ms 22 ms 17 ms 2001:559::b52
8 16 ms 22 ms 16 ms 2607:f8b0:8224::1
9 19 ms 17 ms 17 ms 2001:4860:0:1126::1
10 17 ms 40 ms 18 ms 2001:4860:0:1126::14
11 24 ms 16 ms 17 ms 2001:4860::8:4000:cd80
12 17 ms 35 ms 21 ms 2001:4860::1:0:baf3
13 24 ms 18 ms 17 ms 2001:4860:0:1125::1
14 17 ms 24 ms 17 ms 2001:4860:0:1::1ceb
15 19 ms 18 ms 23 ms lga25s54-in-x0e.1e100.net [2607:f8b0:4006:804::200e]

 

And my pings to google.com for about 5 mins average out at sub 20 with 0 packet loss - which is what I would expect.

Highlighted
New Poster

Re: Latency Variation/Packet Loss, Possible Bufferbloat?

I am experiencing the same issues. Are you located in one of the major cities on the west coast? My guess is the massive increase in traffic due to quarantined citizens is causing bufferbloating at the regional level. My uplink rates are horrendously slow compared to my downlink.