Quote:
Originally Posted by The PIT
There's also a http download cap which operates 24 hrs on a single thread download. So basically although speed test shows say 200mbps you'll between 70mbps and 90 mbps depending on the time of day. Using more than a single download you will reach maximum speed.
Virgin originally said this was by design which is probably true as it's a way of limiting traffic in a oversubscribed network. Then under pressure decided to investigate it. They have since gone rather quiet on the subject. There's 25 page thread on it here https://community.virginmedia.com/t5...026879/page/13
A few people are unaffected I'm not sure whether the gaming bundle has this limit or not.
|
Hi
I'm one of those on the VM forum who is experiencing the issue with single-thread speeds, and have been trying to get them to look into it again (which they are now).
There are a couple of reasons why we believe this is a fault as opposed to a form of traffic management -
- During the day, single-thread download speeds can (rarely) reach full speed, but generally are well below that. They are also extremely unstable and fluctuate constantly anywhere from under 10Mb to over 100Mbps, making the overall average download speed well below the speculated 70Mbps cap.
- At peak time, single thread speeds are severely impacted even for users who don't have a utilisation fault and can otherwise reach full speed on speedtest.net. My area has a long running utilisation fault and although I get about 5-10Mb on speedtest.net, single-thread speeds drop to under 1Mb which makes even general web browsing noticeably slow. This is completely unacceptable and surely no ISP would throttle users so heavily that simple web browsing becomes an issue.
- Everyone who has reported this fault is on a Motorola CMTS. This has been verified by Jen_A on the VM forum who has recently resubmitted this issue to the networks team. She believes this may be a crucial factor which wasn't noted the first time this was raised to Networks. She has found no reports of this problem from users on Cisco or Arris systems.
- Looking through archived speed tests on Think Broadband, this fault first became apparent in November 2015. Before that, single-thread speeds were close to (or matched) the multi-thread speed. ThinkBroadband also noted in their August 2016 speed test article that they are seeing odd single-thread results from parts of the VM network.
There is probably a lot more which I've forgotten to include but they're the main reasons for us believing this is a fault rather than a deliberate traffic management policy.
I did notice some improvement yesterday with a Thinkbroadband test download peaking at 173Mb/s and not wildly fluctuating like normal. Peak time single thread speeds were also vastly improved last night and reaching full speed (10Mbps) so I'm hopeful some progress has been made, but haven't heard anything official yet.