Thanks for the feedback, I’ll stick with it then, but will try testing again and peak times over the holidays to compare.
Virgin Media Discussion Thread
- Thread starter macca40
- Start date
Thanks for the feedback, I’ll stick with it then, but will try testing again and peak times over the holidays to compare.
Soldato
That ain't bad at all, here is my 250/25 connection FQ_Codel over WiFi.... best effort!
Edit: I didn't get chance to finish my testing so... Here is the cake one. This did take me by surprise because it is only using 20% cpu vs FQ_Codel 25/30% usage single core!
@Rainmaker Why does Cake only use a single core, using 'torch' to take a quick look it seems that the test above is multi thread if I looked properly that is ?
Using Arm quad core 64 Mikrotik.
Last edited:
Permabanned
- Joined
- 9 Aug 2008
- Posts
- 35,711
I'm shocked that it doesn't have any advanced features at all. I logged into it and there wasn't much I could configure except port forward (which I'm glad about as I run a micro server that needs external access as it's running web stuff) and also I needed auto static IP's via mac address in the DHCP table.
Does the hardware inside support anything that could be added down the line through software and firmware updates?
Probably wishful thinking!
I secured this for my parents a few months back. Very happy with that.
Now been upgraded to 130mb... no extra charge.
Last edited:
Apparently these are power hungry mofos vs other routers/modems.
Soldato
Are these bufferbloat tests something that Virgin can correct if too high? They seem random when testing back to back for me
That's a question best answered by @Dave Taht if he'd be so kind. I know SQM works, I know it's awesome, and I know enough about bufferbloat to be dangerous... But the inner workings of cake are not something I'm confident on expounding (yet).
Virgin could fix it almost entirely by implementing the DOCSIS 3.1 standard correctly and enabling DOCSIS-PIE at their end; but they haven't, and seemingly won't. You would have to run SQM locally (i.e. on your router) to fix it. The randomness you see (assuming you mean some results being better/worse than others during repeated testing) is because of variability in the network - others using it at the same time in your neighbourhood, other devices on your network, demands on the queue at the NICs etc.
interesting I believe my hub3 was around 15 watts when used more fully, not sure about the hub5 but will test it out soon.
I predict Virgin spending no more of their resources on making any changes to their DOCSIS implementation. Their focus now is overbuilding with fibre and eventually removing the last few metres of DOCSIS as customers churn or upgrade their packages.
Permabanned
- Joined
- 9 Aug 2008
- Posts
- 35,711
Their expansion around the North East is pretty big, can’t imaging the bandwidth needed in total. Blows my mind considering they been thrashing marketing for the 1G.
They probably need a lot less than you think. Most people don't subscribe to the fastest service tiers, and even if they do they don't hammer them 24x7. Even a record evening on TalkTalk only averaged out to 2Mbps per subscriber.
www.ispreview.co.uk
CoD Pushes TalkTalk Broadband Traffic to New 8.71 Tbps Record UPDATE
Internet traffic across UK broadband ISPs surged last night as people tuned in to stream the Champions League semi-final and to download the latest 40GB (GigaBy
www.ispreview.co.uk
Permabanned
- Joined
- 9 Aug 2008
- Posts
- 35,711
They probably need a lot less than you think. Most people don't subscribe to the fastest service tiers, and even if they do they don't hammer them 24x7. Even a record evening on TalkTalk only averaged out to 2Mbps per subscriber.
CoD Pushes TalkTalk Broadband Traffic to New 8.71 Tbps Record UPDATE
Internet traffic across UK broadband ISPs surged last night as people tuned in to stream the Champions League semi-final and to download the latest 40GB (GigaBy
I thought it would have been a lot more bandwidth than this, cool.
Ran quick test, no wifi and modem mode on my hub5 but maxing out the speeds its hitting 12-13 watts, roughly 10 watts on average when not doing anything. Similar if not a bit less then the hub3 really.
This is with an electrical power reader on the wall socket.
A modern SoC was never going to be chewing through power, especially something as barebones as a modem/router. For 10 watts we could be running all manner of cool stuff like an APU or one of the new embedded Ryzen boxes though. Ongoing boo @ Virgin Media for not allowing you to BYOD on the modem side!
Here is mine, 100mb service upgraded to 130mb with a SH5 and PC directly connected it the SH5
www.waveform.com
Bufferbloat and Internet Speed Test - Waveform
View the full results, and test your own bufferbloat
Does 1GB broadband use the BT phone socket?
A friend texted said her 1GB with VM using the BT phone line socket.
I recently ordered 1GB broadband deal via the livechat. They said it has to be their network wall socket, can accept NTL box too. But I don't have any cable wall box in my flat. Just BT wall box, I'm currently with Plusnet.
Is it possible that VM are using different technology?
A friend texted said her 1GB with VM using the BT phone line socket.
I recently ordered 1GB broadband deal via the livechat. They said it has to be their network wall socket, can accept NTL box too. But I don't have any cable wall box in my flat. Just BT wall box, I'm currently with Plusnet.
Is it possible that VM are using different technology?