How do I get lower latency/ping?

[josh94]

The ping has got LOTS better since upgrading to gig1, but it still is high compared to many, many others on here, some on Hub3's... my connection has intermittent low packet loss and the average ping is all over the place.

https://www.thinkbroadband.com/broadband/monitoring/quality/share/06a194e09128090297e5466a95fbc272ef... 

Can someone advise if this is acceptable and I am just being picky on it? My upload is also pathetic 

[josh94]

If I ping say Cloudflare DNS server 1.1.1.1 and do the test it adds ~20ms when downloading and when uploading some lines are 186ms to 1065ms

[josh94]

Interestingly, testing to Voip-Unlimited's server is 13ms ping, which is great... but still half upload....

https://www.speedtest.net/my-result/d/7ae95c79-996e-4356-b067-1e9f0f52ff8f

[legacy1]

---

@josh94 wrote:

What am I pinging?

---

VM DNS its the next closed ping you can do then VM WAN gateway your on but VM DNS IP is easy to understand. 

[Andrew-G · Accepted solution]

---

@josh94 wrote:

Well your saying my levels are good and someone else says they wont be helping, what's your thoughts? I always have high and erratic ping, and my upload fluctuates heavily when no usage is going on, down is not that effected but may just be my areas congested?

---

The levels are crap, trust me.  For starters, the accepted range of DOCSIS power levels on VM's network is up to +10 dBmV, you have multiple channels at or above that.  High power levels are of themselves an indication that the connection is noisy, although within permitted limits that would not matter if they were less than 10, and there were no other symptoms.  Next up, the range of power levels is about 7.3 dBmV from lowest to highest, that's going to cause problems on its own - we frequently see problems when the range exceeds 4.5 dBmV.  Then we have to consider that ideally power levels should be flat or slightly declining as frequency rises - but yours are going up steeply with frequency.  That's an incorrectly set or faulty street cabinet amplifier resulting in incorrect "slope".  There's also uncorrectable errors showing on the DOCSIS 3.1 channel, that's always a bad sign.  Normally I'd expect to also see significant post-RS (uncorrectable) errors on the DOCSIS 3.0 channels.  You'll see values of zero there, but that's not because there aren't any, it's because the Hub 4 firmware is a pile of steaming **** and shows zeroes rather than actual error counts.  I can tell the two columns of zeroes are nonsense because a DOCSIS connection ALWAYS racks up pre-RS errors even if working well.  The pre-RS errors are corrected, they don't matter, but their absence confirms that your hub is running the faulty firmware (like most other Hub 4s).   

So all things considered, you've got a shonky connection, that either an area fault or it's your unique connection, either needs a technician. 

In honesty, VM is not the network for gamers.  When it's working well it's still marginally inferior to the latency of an Openreach connection.  When things go wrong then cable technology is erratic and difficult to troubleshoot.  And then you have VM's world famous technical support.  In return for that, you can download games faster than Openreach, you just can't play them very well.  When you're gaming, it uses less than 10 Mbps up and down.   

[josh94]

Just went into router mode to see if there was any difference to take another test (colchester)... no difference.

Colchester (where I am):

https://www.speedtest.net/my-result/d/f84a0826-d6a2-4c5a-afed-0c32ad6015f3

That IP in router mode not responding to pings either

[legacy1]

---

@josh94 wrote:

If I ping say Cloudflare DNS server 1.1.1.1 and do the test it adds ~20ms when downloading and when uploading some lines are 186ms to 1065ms

---

likely the upstream is congested then

[josh94]

When I had the Hub 3, I had 3 cubic metres of Pre-RS errors and 1 cublic post. I also thought it was hilarious there was none here.

Earlier today I got 1133mbps down and 50.4 up AT MY PC, I was well happy after setting up balance-rr etc... then I looked at my bandwidth graph and did a few more tests and it is chaos.

The RealSpeed shows 847mbps to the Hub 4 and 121mbps to me, 9.7mbps upload to the hub 7mbps upload to me.?????

Speedtest says 1080mbps to me and 21mbps up *more accurate*

iperf3 says 1143mbps to me and 18mbps up *bit more accurate than speedtest over longer test period*

You guys think I should get a technician? To be honest, the wiring is janky as hell, but the connection wasn't like this on docsis3.0. When I purchased this house the wire was visible on the ground and damaged in some areas when the original tech installed it he said 'it tested fine'. I then got VM out again when I had NO UPLOAD during the pandemic <5mbps and although I said you are throttling me he said there was no issues and he left, then he sat in the cabinet which is literally less than 5m from my hub, my connection went off and it came back and I had good upload up until I went to docsis3.1

I also had to ring up yesterday to ask to activate the hub after hours of it not doing it, they said they have sent the signal, an hour later it activated then provisioned on M600. I rung back to be told my ethernet cable is the problem and everything is fine from his end. I hung up and rung back to speak to someone not in the offshore call centre who said there is a work order which is stuck and needed manually processing. He 'processed it' and told me to wait 1 hour.

I wait 3 hours I ring back and say its still M600 they said they can definitely see a work order has processed and I should wait till tomorrow. (yesterday)

I got up this morning and saw it was still 690 max traffic, I spoke to a woman on the phone who said she can see I am only on the M600 plan and that I can upgrade to Gig1 ??????????????????????????? I told her that it is an absolute liability and that I can see Gig1 when I login to my account. She hummed and went off to do some 'checkings' and came back and told me she has free of charge upgraded me to Gig1 and that I can wait 15 minutes and I will have the speed. I told her I absolutely doubt that and gobsmacked I was I restarted and I was provisioned with Gig1.

This all confirms the incompetency of VM.

I suspect I want a tech out who I will show the cable run and he will say 'thats f*d up mate that will need a new cable pulling' and he will also say it needs another team. I don't have a box on the wall either, the house didnt have one, the tech originally (in 2019) said 'it was okay as its protected outside', I knew he was full of s** however I did get my M350 which was the best I could get at that point.

I have a raised wall from the path outside where the box sits next to, its maybe 2.2m x 3m (wide) so literally out my window, the hub is immediately as it comes through the wall. When I exposed the horrific 'crazy paving' to soil the cable was just buried in concrete and soil and even more damaged than I thought, it also has a coupler which was disgusting and looked rusted, I replaced with another coupler thinking that I would need to probably call VM but the line was fine, ping fine, download speed fine, upload speed fine. I guess now using D3.1 there is additional pressure and that this cable is no good.

How do I get someone to come fix things?

Just did a speed test right now in router mode:

 

I am either being throttled or the janky cable from NTL needs to be re-run and a new box on the wall, how do I get this booked in?

[josh94]

VM also text me saying that there is 'essential works' happening on the 1st of Dec to improve my services. I HOPE this is more backhaul as I can almost guarantee my area is congested to f**.

I am putting the hub back into modem mode now... RealSpeed is useless to me, iperf3 says everything

[josh94]

On top of all of this, there is a 'natural spring' according to my water provider which comes out the ground near the cabinet, and out the front of mine through the pavement. They kept digging up around it and trying to find leaks but just said it was a spring each time, smacking the duct in the middle of the pavement with the whacker plate and spades digging under it. There is no way they will pull a cable through that the ducting is...... not good.

It looks like this last winter: 

https://community.virginmedia.com/t5/QuickStart-set-up-and/Frozen-River-down-our-pavement-from-inter...

However they since turned off my neighbours lead supply and coincidently the natural spring has stopped, but the works around the water supply exposed all the ducting and I got to see it, none of it is in good shape.

I am hoping that replacing the cable from my little box out the front of my wall (1.5m from the cabinet) to my hub (including installing a wall plate) will resolve all my issues as I cannot see them doing any work to that ducting including pulling any cables through as you may as well pull fibre through with the condition its in.

I have gone way overboard here with details but I needed to explain it all fully and how f*** everything is, although I did have a good M600 with good up and down with moderately stable ping under usage.

In a couple months I have a FTTH provider being turned on in my street, so I know at the end of this contract I am gone, unless they release the 2.2gig to the public and FTTH doesnt have that, then I will remain lol.

[Andrew-G]

You will have your own reasons for preferring speed over latency, but in your shoes I'd be concerned that VM won't sort out the latency problem.  You have to be prepared to take an Openreach or other ISP connection if need be, because VM's complaint handling is poor and may or may not result in action on the ground, and very occasionally customers encounter "Completely Beyond Us" problems, and if you have a CBU then the company will never fix it, or may take months, even years or persistence on your part to address.

In terms of what your BQM shows, it might be over-utilisation, but with a problematic connection it can be the case that an underlying fault is evident only when the network is under load, rather than being a capacity problem as such.  Which means there's other things going on that need resolving first, and are probably related to or causing the status issues I've picked up - so the grumbling packet loss is an unacceptable problem, and will impact heavily on performance, and the full vertical lines on the BQM indicate either a DOCSIS timeout or other connection interruption, and are often (not always) associated with problems at the cabinet.  I note your comments on a cable re-pull, if that's what's needed, VM need to do it.  Such work is all part and parcel of maintaining buried infrastructure.

Just for reference this is my BQM.  That's not absolutely perfect, but is entirely tolerable for gaming, working from home and other normal uses, as well as being reasonably representative for a decent VM connection.  Note the lack of massive spikes, and the absence of any packet loss.  That's what you should be seeing.

If you are 100% certain that the alternative fibre provider will be available to order in your street, and you would be prepared to take a chance on that, then raise a formal complaint to VM now, asking for them to improve the service problems you are seeing, or to release you from contract without penalty.  Refer to previous problems, but be very clear on what you want fixing now.  Even if VM agree to release without penalty, you don't have to cancel just yet - the purpose of complaining now is so that if the problems persist after the planned network improvements, then you can either immediately exercise the leaving without penalty option, or if they refuse you can use that complaint to take the matter to the adjudication scheme with minimal further delay.

Obviously if the forum staff will take ownership and get this looked at properly then things might happen (how about asking the AFM to become involved, peeps?).  That would be a preferable option for you and the company compared to complaints, escalations and potentially cancelling VM.