Regular Intermittent Upstream dropouts

[SerenayaXIV]

Hi there. I was re-directed here by VM's Twitter. Despite not finding my solution here in the past and instead seeing people with same the issue spending months getting it resolved, I'll humour it and try before making another call to support (my 6th one!)

I am having intermittent upstream issues and have been for about a week now. Every 5-20 minutes or so, usually starting around early afternoon GMT, my upload will crash for seconds at a time. From our contracted 50mb/s to around 10, 12 or less. I've caught this via a SamKnows test but it's difficult to catch in the act since it's brief but very invasive. Why? Because I'm a full-time streamer on Twitch for a career. Unorthodox I know, but I very much rely on upload speeds for this reason. It equates to the bitrate I can send to the platform and therefore stream my gaming as per my job. Phone support does not seem to understand the gravity of the problem this causes. Dropped frames galore at regular intervals. I am on a wired ethernet connection to my router.

I've made several calls to support, only managing to get an engineer after the 4th one. Before that I was being given different answers ranging from "there's a fault in the area" to "it will fix itself within 24 hours". None of this was true of course. I am pretty convinced there is an issue at the exchange somewhere that's not being picked up. I am quite sure others in my area are being affected but they will not notice due to the problem being so intermittent and brief. Most other upload-heavy tasks would not suffer, if they're being used at all.

Anyway, the engineer arrived on Saturday (10th Dec) and could not find the problem. It seems the intermittence is so bad that it eluded him, though his equipment could find no history of errors either. Mind-blowing. I don't really believe this. He tightened some cables, changed a splitter in the box outside and replaced the hub (Hub 4) to be extra sure. No dice of course. No fix. Still a problem. Still can't perform my job amidst a cost-of-living crisis. I am losing my mind. A friend in the industry has advised that it's probably noise/interference and I'd believe it, but where I'm not sure. All I know is it is not on my end. It seems a lot like there is a problem in the area but VM won't budge on checking it since "it's not bad enough", or so the engineer tells me since they'd failed to find a fault at the house even though something is clearly wrong.

I've used a BQM to monitor for days before the engineer visit and it showed nothing out of the ordinary. The only thing I've noticed that correlates with the problem is that one of the channels on the hub drops to 32QAM instead of maintaining 64QAM like the rest when the issue arises. I do not know if this is a cause or effect. Channel 2 specifically. Not Channel ID 2.

Below is the hub as it stands at the time of posting.Right now it looks fine, right? I am aware but sadly this doesn't mean the problem isn't there. Download does not appear to be affected at any point though we are going down to 20% or less of our contracted downloaded speeds at peak times (we're on the 1000/50 contract) which feels pretty awful as well. We are paying a fair amount after all. Engineer also did not find any power issues and the like. However, the upload is far more of a pressing matter right now.

I don't want to hear trivial solutions like checking cables and factory-resetting/restarting a router. It is new and operational and this was all checked literally yesterday. Apologies for the blunt nature of this post, but this is extremely stress-inducing. Please help.

3.0 Upstream channels

Channel Frequency (Hz) Power (dBmV) Symbol Rate (ksps) Modulation Channel ID

1	49600000	43.770599	5120 KSym/sec	64QAM	1
2	23600000	43.020599	5120 KSym/sec	64QAM	5
3	43100000	43.270599	5120 KSym/sec	64QAM	2
4	30100000	43.270599	5120 KSym/sec	64QAM	4
5	36600000	43.270599	5120 KSym/sec	64QAM	3

3.0 Upstream channels

Channel Channel Type T1 Timeouts T2 Timeouts T3 Timeouts T4 Timeouts

1	US_TYPE_STDMA	0	0	0	0
2	US_TYPE_STDMA	0	0	0	0
3	US_TYPE_STDMA	0	0	0	0
4	US_TYPE_STDMA	0	0	0	0
5	US_TYPE_STDMA	0	0	0	0

3.1 Upstream channels

Channel Channel Width (MHz) Power (dBmV) FFT Type Modulation

6

10.8

40.0

2K

QAM8

3.1 Upstream channels

Channel Channel Type Number of Active Subcarriers First Active Subcarrier (Hz) T3 Timeouts T4 Timeouts

6

OFDMA

220

53.4

0

0

[SerenayaXIV]

As an addendum to this, the 6th call I've made has confirmed there is an issue all of a sudden but it's unverified if it's at my house or the area in general as other calls have not come in about it. Again, I don't believe this will happen since Upload is far harder to notice if not under heavy use. Engineer has been booked for the 15th but I still want to know if there is anything else I can do. Is there no possibly way to elevate complaints? Or are we doomed to suffer engineer waiting times until a fault is suddenly found somewhere?

[Sephiroth]

Your problems will occur when the QAM drops. What happens is that the symbol rate on that channel halves when the VM end detects noise. This slows all the upstream down because the real data packets have to be assembled on arrival and that will be delayed by the speed of the noisy channel.

If your upstream is behaving like that, so will the upstream for up to at least 500 other users in your locality depending on where the noise arises. So 500 users are piling in on an impaired upstream, contention rises exponentially (queuing theory) and that’s what you’re seeing.

The VM bods on this forum must escalate this to the Network Team and NOT tell you that at the moment they looked at your modem, all was in spec.

Keep us posted, please.

[SerenayaXIV]

Thank you for the response. Is there any way to alert customer support on the phone of this and ask for them to escalate to networking team with that evidence? Or do I have to hope that forum mods notice this thread?

[Sephiroth]

Customer Support on the phone are total ignoramuses, imo. This needs escalation to Networks Team. The problem here is that the QAM variations are what the system is supposed to do when there’s noise. But persistent noise is NOT normal and the Ndetwoprk Team would have to deal with this.

You have two options, I’d say:

1/
Hack your way with the Forum Team (who are not ignoramuses, but nor do they know the ins and outs of the DOCSIS protocol).

2/
Call RETENTIONS and hack it out with them. They are generally reckoned to be reasonable people with some heft.

[SerenayaXIV]

I hope the forum team notices the thread then. VM's Twitter at least seemed confident they could help but of course they would be. No response yet either way. Not sure how one would get their attention outside of hoping they see this.

I tried with customer service one last time just now. They were unwilling to budge and they're adamant they have no contact with the network team. Engineers are the only people that can but I have to wait 4 more days on top of the week I've already spent at a stand-still. Time is certainly money so, that's a pretty stressful wait given that the first engineer didn't even want to try contacting them when I'd brought it up due to reasons given in my original post.

Retentions seems risky since...well, as much as I'd love to not deal with this lack of stable service, we don't have many other options here. A potential price reduction also won't fix the issue. VM seem to monopolise the entire area and good fibre is nowhere to be seen. If the forum team cannot help, my only real hope is that there's apparently work due soon in the area to increase the upload cap, which may either fix the noise problem or make it matter less. Engineer could not give a date for that, though. Of course...

Here's to hoping.

[SerenayaXIV]

3.0 Upstream channels

Channel Frequency (Hz) Power (dBmV) Symbol Rate (ksps) Modulation Channel ID

1	49600000	43.770599	5120 KSym/sec	64QAM	1
2	23600000	43.270599	5120 KSym/sec	32QAM	5
3	43100000	43.520599	5120 KSym/sec	64QAM	2
4	30100000	43.270599	5120 KSym/sec	64QAM	4
5	36600000	43.520599	5120 KSym/sec	64QAM	3

After a period of some stability today, I just had a huge drop when testing frame drops with OBS (the only reliable way for me to see it happening in real time since speed tests like SamKnows requires me sitting there pressing "Launch" over and over). As per usual, the same channel has dropped to 32QAM at the same time. Please can a forum team member take a look at this thread and escalate this? I've been dealing with this for a week now and it's extremely detrimental to my livelihood. An engineer couldn't find the problem and customer support have not been helpful. All other avenues have proved fruitless thus far. Thank you.

[jbrennand]

I have asked them to take a look and respond here

[Nathan_B]

Hi there @SerenayaXIV, welcome to our forum and thanks for your post.

I'm sorry to see you have been having issues with your connection dropping out. I can see that you have spoken with the team since posting this. Were they able to help with getting this resolved.

If not, please let us know so we can assist further.

Regards

Nathan

[SerenayaXIV]

Hi there, Nathan.

No, they could not. The problem described above is very much still present. The most I have is another engineer booked for Thursday (15th) but honestly after a week of this upload issue, I don't particularly want to wait that long just to be told the same thing the first engineer (yesterday) told me as per my original post. They will not find the problem just like the first one, said everything was fine in and around my house and they won't refer it to network engineers without the service visibly affecting other people (or them reporting it themselves). Or so they say. I am very sure it is affecting other households, but they won't notice due to its intermittence and them not really relying on upload specifically. It's not like it's download speeds or complete broadband dropouts after all. I understand my particular occupation is not common, but that's why it's an issue. I rely on its stability for work.

Customer support on the phone give me different answers every time and refuse to take the complaint any higher. Networking Team seem uncontactable from the customers end to report potential area-wide faults.

Is there anything you're able to do to help?

Thanks.