Virgin Media Community

[FlyingZebra]

Hi all

Hoping somebody can help with this issue. Long time VM user and lurker but first time poster.

I have recently upgraded from M350 to M600 broadband as part of the Ultimate Oomph bundle. I have previously had some issues with minor packet loss and higher latency & jitter on the line, but it had got better lately. On the install date (06/05), the line appeared to get worse following the install appointment - although I assumed this was normal setup process for the new profile and it would work out eventually (similar to other networks' negotiation process). The line had looked to have settled down in the evening of 08/05 and there were no practical issues over the weekend with calls and the like. The line remained fully stable with a low ping and low jitter until 11/05 - the BQMs were showing pretty much flat.

However, the issue now reoccured yesterday evening (11/05), causing the latency to spike right up again with occasional dropped packets as well.

The Hub is used primarily in modem mode (although the issue appears to occur in router mode too). There has not been any significant load from my side on the network, at least that has changed during the time periods in question. There are normally background uploads for backup but again this has been constant throughout and has not changed.

It should not be any internal equipment either as I have a separate line (via Openreach) that is solid as a rock and is connected to the same place with Ethernet. I would expect to see both lines having similar issues if it was an internal issue. Also when using the Hub in router mode the same is observed. Virgin support over the phone cannot seem to find any resolvable issues. I have also tested using manual ping against a variety of sources in case of upstream issues and that replicates the issue as well.

BQMs and status data attached below, any guidance would be much appreciated.

[FlyingZebra]

Downstream bonded channels
Channel Frequency (Hz) Power (dBmV) SNR (dB) Modulation Channel ID
1 331000000 3 40 256 qam 25
2 203000000 2.4 38 256 qam 9
3 211000000 2.4 38 256 qam 10
4 219000000 2.4 40 256 qam 11
5 227000000 2.5 40 256 qam 12
6 235000000 2.5 40 256 qam 13
7 243000000 2.4 38 256 qam 14
8 251000000 2.2 40 256 qam 15
9 259000000 2.2 38 256 qam 16
10 267000000 2.2 40 256 qam 17
11 275000000 2.2 40 256 qam 18
12 283000000 2.2 40 256 qam 19
13 291000000 2.5 40 256 qam 20
14 299000000 2.7 40 256 qam 21
15 307000000 2.5 40 256 qam 22
16 315000000 2.7 40 256 qam 23
17 323000000 3.2 40 256 qam 24
18 339000000 3.4 40 256 qam 26
19 347000000 2.9 40 256 qam 27
20 355000000 2.7 40 256 qam 28
21 363000000 2.5 40 256 qam 29
22 371000000 2.5 40 256 qam 30
23 379000000 2.5 40 256 qam 31
24 387000000 2.7 40 256 qam 32

Downstream bonded channels
Channel Locked Status RxMER (dB) Pre RS Errors Post RS Errors
1 Locked 40.3 6 0
2 Locked 38.9 6 0
3 Locked 38.9 5 0
4 Locked 40.3 4 0
5 Locked 40.3 5 0
6 Locked 40.3 0 0
7 Locked 38.9 0 0
8 Locked 40.3 0 0
9 Locked 38.9 6 0
10 Locked 40.3 0 0
11 Locked 40.3 6 0
12 Locked 40.3 0 0
13 Locked 40.3 5 0
14 Locked 40.3 6 0
15 Locked 40.3 3 0
16 Locked 40.3 0 0
17 Locked 40.3 5 0
18 Locked 40.3 7 0
19 Locked 40.3 0 0
20 Locked 40.3 0 0
21 Locked 40.3 5 0
22 Locked 40.3 6 0
23 Locked 40.3 4 0
24 Locked 40.3 0 0

Upstream bonded channels

Channel Frequency (Hz) Power (dBmV) Symbol Rate (ksps) Modulation Channel ID
1 39400000 48.5 5120 64 qam 4
2 46200000 49 5120 64 qam 3
3 32600000 48.3 5120 64 qam 5
4 25800000 48 5120 64 qam 6

Upstream bonded channels

Channel Channel Type T1 Timeouts T2 Timeouts T3 Timeouts T4 Timeouts
1 ATDMA 0 0 0 0
2 ATDMA 0 0 0 0
3 ATDMA 0 0 0 0
4 ATDMA 0 0 0 0

General Configuration
Network access
Allowed
Maximum Number of CPEs
1
Baseline Privacy
Enabled
DOCSIS Mode
Docsis30
Config file
REMOVED

Primary Downstream Service Flow
SFID 908742
Max Traffic Rate 690000278
Max Traffic Burst 42600
Min Traffic Rate 0

Primary Upstream Service Flow
SFID 908741
Max Traffic Rate 44000278
Max Traffic Burst 42600
Min Traffic Rate 0
Max Concatenated Burst 42600
Scheduling Type BestEffort

10/05/2021 19:42:18 Error DHCP RENEW WARNING - Field invalid in response v4 option;CM-MAC=DELETED;CMTS-MAC=DELETED;CM-QOS=1.1;CM-VER=3.0;

[FlyingZebra]

I have now done some further investigation and it appears to be something to do with the Hub 3 that I have not coping well at a high percentage upload/download speed. It is most apparent when saturating the upload speed which without anything in the way will pretty much without fail cause dropped ping packets at the same time. I have attached below a BQM for my other line connected to the router (with the Virgin Hub in modem mode at the same time) suggesting that my router is handing packets without an issue, pointing me to the Hub itself.

As a temporary resolution, I have set an artificially low bandwidth rate for the 'smart queues' / fq_codel feature on my own router. This appears to help the issue but it's not great having to hobble the bandwidth to prevent latency issues.

I am awaiting an engineer to check the line and I will see how that goes.

[legacy1]

---

@FlyingZebra wrote:

As a temporary resolution, I have set an artificially low bandwidth rate for the 'smart queues' / fq_codel feature on my own router. This appears to help the issue but it's not great having to hobble the bandwidth to prevent latency issues.

---

But thats the what you have to do QoS the the line I have to limit my upload to 9.1Mb plus download to stop ping spikes or jitter as the hub don't QoS your connection its just a single queue and you buffer due to the queue buffer being to big and packets get delayed the faster you go the less of a problem it is but upstream can be a bigger problem if the available bandwidth is not their you just buffer at the hub. 

[FlyingZebra]

Thank you legacy1 for your advice here. I have now set the maximum bandwidth at 400/35 with the smart queue / fq_codel option. From testing, this is the most I can get away with whilst keeping a semi-reasonable line latency and reduced packet drops. There still appear to be irregular spikes in the latency no matter what I do, but I am not sure if they can be fully solved or just mitigated. I will continue to monitor on the BQM.

I have had the engineer out who has changed some old plugs and also assisted in testing with a different Hub 3. However the problem still persists with both options, even with the Hub 3 in modem mode and connected via ethernet, which leads me to either the local network or the Hub 3. The engineer had no further suggestions.

DOCSIS 3.1 may help in this situation, as the spec apparently includes 'active queue management' on the modem which may assist, but it is not rolled out yet in my city. I have tried asking customer services for a Hub 4, but I am told they are only for Gig1 customers at this time.

Do I have any other options here? It seems that if I want to ever actually use the full 600mbps of my new plan, I have to significantly sacrifice latency, jitter and packet loss rates, causing issues with games and phone calls.

[legacy1]

You should be able to go higher for the upload to 39Mb and 590Mb for download....unless and I don't have a hub 3 only 100M the rate limiting starts to delay packets egressing the closer you get to the set limit, which would mean you have to limit more for the hub not to delay packets.

What I do is ping the VM WAN gateway IP which is about 5-10ms then do a file download or upload to see how much the ping jump to. You ideally have to do this at low peak times.

[FlyingZebra]

This is the issue, anything above around 400/35 starts to cause packet loss and latency spikes and jitter. Even the 400/35 setting shows  this, but it is more managable.

I run similar tests - I use regular pings (either to Cloudflare's 1.1.1.1 or one of my servers) whilst doing a speedtest. I also find the Waveform Bufferbloat test to be useful in testing loaded latency. The only issue is that at higher than the above bandwidths, the latency gets significantly worse.