Newsgroups Managed

[dammydam]

Hi, All
I've just browsed over the traffic management policy at Virginmedia. It seems there has been a slight update to the policy regarding Newsgroups and P2P.

Here's a copy and paste of the section i am referring to.

File Sharing:
We also moderate the total volume of file sharing traffic on our network between 5pm and midnight on weekdays and midday and midnight on weekends. This policy is restricted to Peer to Peer (“P2P”) applications and Newsgroups (which are commonly used to distribute large amounts of data)

This policy does not impact any applications other than Peer to Peer and Newsgroups, so things like watching iPlayer, online gaming, making calls via Skype, downloading music tracks from iTunes or streaming them from Spotify and sending an email or normal browsing are unaffected.

It's important to remember that these traffic management policies only apply at peak times when speeds are most likely to be affected by people using more than their fair share. Outside of peak times we do not manage traffic.

It there any informed forum members who could shed any light on what restrictions are now in place regarding data allowances? .

Sorry if this issue has already been discussed

David

[weesteev]

Looks to me like Traffic management limitations only apply to these sources of data transfer... sounds good. Get those leachers capped!

[dammydam]

Looks to me like Traffic management limitations only apply to these sources of data transfer... sounds good. Get those leachers capped!

Do you really represent Virginmedia? I find it quite alarming that you refer to your customer base as Leechers. I should imagine some people reading this would be offended by your comments.

Any ways my question was about how the shaping policy is implemented. Will the shaping rules restrict everyone between the times stated. Or will it work based on your own personal usage?

David

[Pantsu-san]

As World of Warcraft patches, which are generally large in size, use P2P distribution does this mean that I can expect to wait days instead of hours for the large updates? I've already had to download a updated 15GB beta client and the incoming 4.01 patch is looking like it will be double figures in GB.

Would love someone who works for Virgin (and doesn't consider me a leecher) to inform me of where we stand in situations like this. Thanks.

[Monkeylord]

"Leecher" is actually the term used for P2P downloaders who do not yet have the whole download, and is either just downloading or both downloading and uploading at the same time. Someone who does have the whole download and is sharing it without downloading is known as a "Seeder".

So, as you can see, "Leecher" isn't a derogatory term.

[Hugh]

As World of Warcraft patches, which are generally large in size, use P2P distribution does this mean that I can expect to wait days instead of hours for the large updates? I've already had to download a updated 15GB beta client and the incoming 4.01 patch is looking like it will be double figures in GB.

Would love someone who works for Virgin (and doesn't consider me a leecher) to inform me of where we stand in situations like this. Thanks.
You could always download it outside the STM periods.

[broadbandking]

You could always download it outside the STM periods.

That's against his human rights.:D

[Pantsu-san]

I'll be downloading the WOW updates as soon as they're published no matter if it's inside or outside of STM periods. I was enquiring if they'd also be hit with the 'unlimited unless we say so' tax. Paying for the 50Mb, no STM, service now appears to be a redundant choice for me.

That's against his human rights.:D

Your sheer existence should be against my human rights.

[broadbandking]

Now now Pantsu-san was there any need for that comment, I was merely adding a light spin on the situation hence the smile

[weesteev]

Do you really represent Virginmedia? I find it quite alarming that you refer to your customer base as Leechers. I should imagine some people reading this would be offended by your comments.

Any ways my question was about how the shaping policy is implemented. Will the shaping rules restrict everyone between the times stated. Or will it work based on your own personal usage?

David

I dont represent VirginMedia, my signature makes that very clear. It is my own personal view.

I do download a hell of a lot of data myself but the arrogance of some downloaders as astounding, the same people who then have the nerve to complain about slow speeds when they want to downlaod when they generally leave 20 Torrents seeding all day and night!

If download limits are a concern then its worth looking at an unrestricted service like 50Mb.

Shaping limits will not come into effect on an entire node, any restriction will be assigned to usage by MAC address through the CMTS. It will be extremely controversial to restrict every local user when a few of them hit their limit, although I have heard of that on other suppliers before.

[Ignitionnet]

I dont represent VirginMedia, my signature makes that very clear. It is my own personal view.

I do download a hell of a lot of data myself but the arrogance of some downloaders as astounding, the same people who then have the nerve to complain about slow speeds when they want to downlaod when they generally leave 20 Torrents seeding all day and night!

If download limits are a concern then its worth looking at an unrestricted service like 50Mb.

Shaping limits will not come into effect on an entire node, any restriction will be assigned to usage by MAC address through the CMTS. It will be extremely controversial to restrict every local user when a few of them hit their limit, although I have heard of that on other suppliers before.

The shaping will be applied by protocol, not by user MAC address. that remains the province of STM.

It's not remotely controversial to do this, every supplier implementing protocol based management applies it across the entire user base with different profiles applied to different users perhaps but nonetheless the point of this isn't to affect customers who are overusing, it's to restrain the % of network resources being soaked up by certain non-time critical protocols.

[Pantsu-san]

I was merely adding a light spin on the situation hence the smile

I was also smiling when writing. Does that count?

---------- Post added at 15:15 ---------- Previous post was at 15:09 ----------

If download limits are a concern then its worth looking at an unrestricted service like 50Mb.

Are you suggesting that the 50Mb service will only suffer from STM when uploading over 6GB? Is that the case? If so, then my worries about downloading the WOW patches are unwarranted as they're rarely this big - only when a new expansion rolls around.

[weesteev]

50Mb doesnt have a download limit, only a new Upload limit. I would be amazed if your WOW patches breached your upload limit! My upload rate is always very low when patching WOW.

[Ignitionnet]

Per another thread shaping and STM are not related, they will run side by side. Shaping active on all tiers between 5pm and 12 midnight, STM per the table.

[weesteev]

The shaping relates to a 25% network traffic figure, do you think that will be per Node or across the entire network? If its per node then some areas (especially student areas) are going to get massacred!

Not a bad thing though i suppose ;)

[Pantsu-san]

Per another thread shaping and STM are not related, they will run side by side. Shaping active on all tiers between 5pm and 12 midnight, STM per the table.

That is much clearer. Thanks, Igni.

So even being on 50Mb isn't going to matter. Large WOW downloads are going to take longer as I inititally suspected along with any other beta's that I am involved with that use a P2P method to patch/deliver the client. That's not pleasing to my ears.

I need to enquire with Blizzard if the new streaming beta client still uses P2P. The rapid download of the beta patches seem to be too rapid to suggest it does but as it's a beta, who knows what method of delivery will be in the live client.

Thanks.

[Ignitionnet]

The shaping relates to a 25% network traffic figure, do you think that will be per Node or across the entire network? If its per node then some areas (especially student areas) are going to get massacred!

Not a bad thing though i suppose ;)

I believe traffic will be identified by DPI kit and tagged at the edge of the network then dropped as required via regular QoS.

Will happen per logical interface, so each and every cable interface can be managed.

This is a big 'As far as I know' though and it doesn't fit in with upstream shaping being enabled.

[zantarous]

S what exactly is the policy for newsgroups as I either do my downloading before 10am but as I am now going to be doing a 9-5 job again will be doing it after 9pm. Is it a straight throttle or after x amount?

[RainmakerRaw]

As per my post here (http://www.cableforum.co.uk/board/35100999-post47.html) VM have confirmed on their site that NNTP and BT will be throttled on XXL/50Mbps. Exactly what "25% of capacity" means though, who knows? Does that mean if I'm the only downloader in my area at a given time, that I'll have 25% of a metaphorical gazillion Mbps ("the capacity") and hence my full 50 megs as per? Or 25% of my 50Mbps? Or what? It's as clear as a dirty pond.

Does this mean all SSL traffic is to be shaped, including VPN/work connections, financial transactions etc? Even with DPI I'm not sure how they'll be able to tell apart NNTP from other secured data on a given port (119, 443, 563 and so on). Anyone?

[Ignitionnet]

As per my post here (http://www.cableforum.co.uk/board/35100999-post47.html) VM have confirmed on their site that NNTP and BT will be throttled on XXL/50Mbps. Exactly what "25% of capacity" means though, who knows? Does that mean if I'm the only downloader in my area at a given time, that I'll have 25% of a metaphorical gazillion Mbps ("the capacity") and hence my full 50 megs as per? Or 25% of my 50Mbps? Or what? It's as clear as a dirty pond.

Does this mean all SSL traffic is to be shaped, including VPN/work connections, financial transactions etc? Even with DPI I'm not sure how they'll be able to tell apart NNTP from other secured data on a given port (119, 443, 563 and so on). Anyone?

You don't need to know exactly what the 10 SSL connections to news.giganews.com are carrying to know they're going to be NNTP.

25% of capacity means 25% of any given network interface.

---------- Post added at 17:51 ---------- Previous post was at 17:50 ----------

S what exactly is the policy for newsgroups as I either do my downloading before 10am but as I am now going to be doing a 9-5 job again will be doing it after 9pm. Is it a straight throttle or after x amount?

Straight throttle, variable depending on network load.

[Sirius]

You don't need to know exactly what the 10 SSL connections to news.giganews.com are carrying to know they're going to be NNTP.

25% of capacity means 25% of any given network interface.

---------- Post added at 17:51 ---------- Previous post was at 17:50 ----------



Straight throttle, variable depending on network load.Cannot wait to see if my ssl and vpn will be affected by it :)

[RainmakerRaw]

You don't need to know exactly what the 10 SSL connections to news.giganews.com are carrying to know they're going to be NNTP.

As per the other thread, that would mean a trivial privately owned and operated NNTP proxy set up on a user's dedi-server (not known to any ISP as a Usenet provider) or else using a decent VPN. The point being if someone wants to find a way around it, they likely will.

25% of capacity means 25% of any given network interface.

Genuine question, but what does that mean in plain English/layman's terms? Does that mean we 'will' be getting less than 50Mbps even when the system isn't being otherwise utilised? Or does it mean we can still get line max if others aren't wanting the bandwidth at the same time?

As with anything, it's a case of seeing how it goes I suppose. I'm nocturnal so this won't really affect me btw. I'm not discussing this because I'm some super-angry mega-leecher. I'm a disabled guy who is forced to all but live in his bed, who relies on the internet for, well, EVERYTHING that doesn't involve sleeping.

"Socialising" via forums and IM, TV shows (don't need a TV licence if it's not live broadcasts, which is a plus), communication, virtually everything. So restrictions on my service are always frowned upon and I'm happy to vote with my wallet. But as I said, I'm mostly up at nights anyway so it's more principle than "OMG you killed my uber downloads?!1!". :)

[Ignitionnet]

Cannot wait to see if my ssl and vpn will be affected by it :)

SSL to the servers using the freebie included SSL will be yes :)

---------- Post added at 18:00 ---------- Previous post was at 17:58 ----------

As per the other thread, that would mean a trivial privately owned and operated NNTP proxy set up on a user's dedi-server (not known to any ISP as a Usenet provider) or else using a decent VPN. The point being if someone wants to find a way around it, they likely will.

I absolutely agree, however the vast majority won't be clueful enough and will deal with it. Nothing you can do about the minority who will spend time and money on such things.

Genuine question, but what does that mean in plain English/layman's terms? Does that mean we 'will' be getting less than 50Mbps even when the system isn't being otherwise utilised? Or does it mean we can still get line max if others aren't wanting the bandwidth at the same time?

As I read it P2P/NNTP/Etc will get at a maximum 25% of network resource, so it entirely depends on how many other people you share resources with are downloading. If that 25% of the port gives you line rate you get line rate, otherwise you share with others.

[Sirius]

SSL to the servers using the freebie included SSL will be yes :)

---------- Post added at 18:00 ---------- Previous post was at 17:58 ----------



I absolutely agree, however the vast majority won't be clueful enough and will deal with it. Nothing you can do about the minority who will spend time and money on such things.



As I read it P2P/NNTP/Etc will get at a maximum 25% of network resource, so it entirely depends on how many other people you share resources with are downloading. If that 25% of the port gives you line rate you get line rate, otherwise you share with others.

would be no

company in france for my vpn :)

Used only when i download and i send that traffic through vpn :)

I have had my hand forced by the likes of detica

[Ignitionnet]

Should do the throttling for VM in that case, those companies running such things usually have inadequate bandwidth and services run like crap. :)

If you've an exception to the rule nicely done. The point is that most won't so it's all good.

[Sirius]

Should do the throttling for VM in that case, those companies running such things usually have inadequate bandwidth and services run like crap. :)

If you've an exception to the rule nicely done. The point is that most won't so it's all good.

As i said i was going to do something to protect myself from the big D anyway

[Chrysalis]

The shaping will be applied by protocol, not by user MAC address. that remains the province of STM.

It's not remotely controversial to do this, every supplier implementing protocol based management applies it across the entire user base with different profiles applied to different users perhaps but nonetheless the point of this isn't to affect customers who are overusing, it's to restrain the % of network resources being soaked up by certain non-time critical protocols.

since you quote left wing sometimes, isnt traffic shaping a sort of left wing policy? ensuring bandwidth is spread out to the protocols that arent agressive.

What is and what isnt time sensitive can only be decided by the end user, if I start a download and want/need it to be fast ie. I am sitting waiting for it then isnt it time sensitive?

It will be interesting to see if VM follow the steps of other isps and we see completely over the top anal throttling with reports of p2p speeds under isdn speeds.

I dont use p2p or nttp except rarely if something I need is only on p2p, so in theory this will benefit me, but my past experience of protocol shaping is it gets things wrong aka false positives and plus I am against things that break net neutrality. I ahve also noticed it tends to not actually fix isp related speed issues, look at O2 access service, heavily throttled yet very congested still, and how plusnet was before it decided to actually buy more capacity. Protocol shaping seems to be more used as a substitute for capacity rather than thinking about end users.

[SideWeaver]

It was always going to happen. They have got to look after their network at the end of the day.

[RainmakerRaw]

Exactly. As I said, I'd rather pay double and not have a crippled connection.

[watzizname]

It was always going to happen. They have got to look after their network at the end of the day.

Right, and fill it to the brim with even more subscribers.

[Ignitionnet]

since you quote left wing sometimes, isnt traffic shaping a sort of left wing policy? ensuring bandwidth is spread out to the protocols that arent agressive.

No it's quite the opposite, enforcing some measure of moderation on usage so that people can't claim all of a shared resource but instead receive closer to what they pay for rather than what other people pay for as well. Getting out what you put in rather than what you put in and what other people put in is certainly not left wing. It's not spreading a resource it's preventing disproportionate usage of the resource by a minority.

Interesting approach, making shaping political :)

---------- Post added at 23:51 ---------- Previous post was at 23:50 ----------

What is and what isnt time sensitive can only be decided by the end user, if I start a download and want/need it to be fast ie. I am sitting waiting for it then isnt it time sensitive?

The protocol itself isn't time sensitive. Streaming if the bandwidth is not available simply breaks. Downloads are not a real time operation.

[dan dority]

ive only had xxl 50mb for 6 weeks and now virgin are going to start traffic shaping it.
robbers :mad:

[colin25]

Perhaps disingenuous, they are not stealing..just not giving :)

[Chrysalis]

The protocol itself isn't time sensitive. Streaming if the bandwidth is not available simply breaks. Downloads are not a real time operation.

ok why are the following considered time sensitive? they dont break if congested.

http
email - this one especially

[Ignitionnet]

ok why are the following considered time sensitive? they dont break if congested.

http
email - this one especially

HTTP can carry time sensitive traffic and is an interactive operation, people click to follow a link and expect a response in a timely fashion and in some instances even are using an interactive real-time application, email carries a minimal traffic volume and the overheads of policing it aren't worth it relative to this.

I have configured shaping policies in the past, so I'm more than happy to play this ;)

[Chrysalis]

your response on http is similiar to what I expected.

I am not playing any games just asking genuine questions. So your answer was because people expect a timely response it is considered interactive. a bit different to saying something is interactive because it breaks if packets are dropped or delayed. So if I were to download something over p2p and expect to not have to wait long why is it not interactive? eg. I may choose to download a driver package (yes there is a guy who makes drivers and only distributes over p2p) and I may be needing to wait for these to download to continue with my work, that makes it an interactive operation.

I understand what you said about email but its a bit different to saying email is time sensitive as isp's do, what you said is the truth, its not worth throttling as the traffic volume is too low.

[Ignitionnet]

So if I were to download something over p2p and expect to not have to wait long why is it not interactive? eg. I may choose to download a driver package (yes there is a guy who makes drivers and only distributes over p2p) and I may be needing to wait for these to download to continue with my work, that makes it an interactive operation.

Downloads are not interactive, a person doesn't hit the download button and expect an immediate completion they hit it and have to wait for it to complete.

HTTP is an attended, interactive operation. Nothing happens without a human being being present and continuation depends on a response. Downloading does not depend on constant human interaction, it is a click then an unattended operation.

Whether you need the download to continue your work or not isn't the issue, the download itself does not require any interaction once it's been initiated and downloading another file does not depend on the previous file while web browsing most certainly depends on previous interaction - one cannot follow a link without having the previous page loaded. HTTP is an interactive application whose experience is dependent on ongoing responsiveness, you may want that download qiuckly however it being slowed does not affect the end result - you'll get the file just more slowly.

The only exception to this would be using P2P to download another P2P downloader, doesn't happen often.

Next please.

[sam1989]

personally i can just see this as a measure to get people to upgrade to 100mb when it comes and they will advertise that as truly unlimited just as 50mb was when i signed up .

[Ignitionnet]

personally i can just see this as a measure to get people to upgrade to 100mb when it comes and they will advertise that as truly unlimited just as 50mb was when i signed up .

All tiers will be shaped.

[Chrysalis]

Downloads are not interactive, a person doesn't hit the download button and expect an immediate completion they hit it and have to wait for it to complete.

HTTP is an attended, interactive operation. Nothing happens without a human being being present and continuation depends on a response. Downloading does not depend on constant human interaction, it is a click then an unattended operation.

Whether you need the download to continue your work or not isn't the issue, the download itself does not require any interaction once it's been initiated and downloading another file does not depend on the previous file while web browsing most certainly depends on previous interaction - one cannot follow a link without having the previous page loaded. HTTP is an interactive application whose experience is dependent on ongoing responsiveness, you may want that download qiuckly however it being slowed does not affect the end result - you'll get the file just more slowly.

The only exception to this would be using P2P to download another P2P downloader, doesn't happen often.

Next please.

ignition you have completely lost me sorry.

if I start a download it is interactive full stop, if I am waiting for that download to finish before I can do something else it is time sensitive. Yes some downloads are not time sensitive in the sense you dont mind leaving them going in the background but some are in that you need them to finish to carry on with what you are doing. This is what makes protocol shaping so wrong, an isp cannot decide for all their customers what is urgent and not urgent.

Do you think I pay for 20mbit over 10mbit so I can browse a http site faster?, for that 2mbit is enough. There is technically nothing different in the amount of clicks between starting a download and starting a http page load. A weak response to try and justify a policy which robs pete to pay paul.

You also avoided the question I asked in my other post as well in regards as to why do they need to implement this if as you claim they have no capacity issues in most areas. Something which you seem to know a lot about for someone who doesnt work for VM anymore.

In short I can understand for things like VOIP, VOD, gaming, ssh. However email and http are no more time sensitive then downloading.

[caph]

I wish the CPS would just instigate a case against a random handful of newsgroup subscribers downloading illegal material. A couple of suspended jail terms handed out would I think have an incredible effect in reducing bandwidth usage on the VM network and would probably go a long way to reducing the need for STM/shaping in the first place.

[Chrysalis]

I wish the CPS would just instigate a case against a random handful of newsgroup subscribers downloading illegal material. A couple of suspended jail terms handed out would I think have an incredible effect in reducing bandwidth usage on the VM network and would probably go a long way to reducing the need for STM/shaping in the first place.

jail terms get handed out all the time it doesnt stop people, the question if they made a big publicity thing of it would it stop people? more likely it would lose VM customers as I dont think they would want potential customers think the isp is involved in arrests.

Its not a criminal offense to download copyright material which is all people are doing from newsgroups.

[Ignitionnet]

if I start a download it is interactive full stop, if I am waiting for that download to finish before I can do something else it is time sensitive. Yes some downloads are not time sensitive in the sense you dont mind leaving them going in the background but some are in that you need them to finish to carry on with what you are doing. This is what makes protocol shaping so wrong, an isp cannot decide for all their customers what is urgent and not urgent.

Of course it isn't. Go check the definition of interactive. The link you click on in the web page rather depends on the web page loading.

It's perfectly appropriate for the ISP to have a policy that will not prejudice the majority of its' customers. How large is a single driver and what is the impact of that being slowed to 50% or less of 20Mbit? Chances are the file if being delivered over P2P isn't going to be that much faster anyway.

Compare that to the people who are downloading a ton of data, largely at peak times, using these protocols.

Do you think I pay for 20mbit over 10mbit so I can browse a http site faster?, for that 2mbit is enough. There is technically nothing different in the amount of clicks between starting a download and starting a http page load. A weak response to try and justify a policy which robs pete to pay paul.

So you click one HTTP link in isolation always. You're not, say, clicking a link, then perhaps following a link in that page, etc? As in waiting on feedback from one action before using that for the next one?

You also avoided the question I asked in my other post as well in regards as to why do they need to implement this if as you claim they have no capacity issues in most areas. Something which you seem to know a lot about for someone who doesnt work for VM anymore.

Streaming media usage is going up extremely quickly. In some pockets however it's drowned out by P2P and Usenet. As I'm sure you're aware VM can't just keep throwing capacity at the issues. It's pretty evident they are upgrading for the new speeds else they'd just shape the hell out of everything 24x7 and release them. It's also pretty evident there aren't network wide capacity issues else there would be many more complaints than there are, most of the issues reported here and on official forum are signal, not capacity related.

I also mentioned statistical contention, this becomes an 'issue' on 100M again.

In short I can understand for things like VOIP, VOD, gaming, ssh. However email and http are no more time sensitive then downloading.

So streaming video carried over HTTP is no more time sensitive than downloading? Or browsing the web, as one almost never does a single HTTP transaction in isolation but a series of them depending on others. Slow each page by 50% and after 4 pages you've just increased the time taken to traverse those by a factor of 8. It's cumulative unlike a download which is almost never conducive to further delay, the one exception being when downloading a downloader for something else.

Again downloading is not interactive, you can start a download and make a cup of tea, try doing that while clicking through links on Google et al searching for something. One activity is interactive, depends on reponses to previous inputs, the other isn't.

Email can be real time, people do in the odd occasion use it to communicate, and regardless traffic volume is minimal.

Lowering the priority of any protocol will disrupt some people unfortunately, it's never an ideal solution however these are early days, the policy will no doubt be subject to change and tuning. That is something that's great about this stuff, it's so flexible.

Wait and see is probably as good a way to roll as any right now.

[Chrysalis]

true I can make a cup of tea whilst downloading, however if the throttling is over the top (like most isps do) and goes at isdn speeds, I can probably drive down to london for a day trip and come back before its finished :)

streaming over http obviously is time sensitive but rather I meant just loading something like a web article. It is time sensitive in the manner you need to wait for it to load so can read the page, but that is the same as you cannot use what you download until it is downloaded.

The driver I am reffering to in this case is driverpacks for windows and also packages used on customised installation discs (all legal) some can be quite small and some can be a couple hundred meg. I tend to download them only when I need them such as when installing windows on some machines and wanting the most up to date integrated stuff. Granted the last time I did this was probably 2008 but like you said I will wait until I dish out proper judgement.

I guess what you are reffering to is like what I had on legacy VM, browsing on that was noticebly slow yes I wont lie, it was inconveniant but was useable. The real problem then was the lack of ability to use streaming sites, watch F1 online etc. and hideous download speeds. Waiting an extra couple of secs for a page is nothing compared to other problems.

If it was a simple prioritisation system rather than this max 25% of capacity put out, then that would be more reasonable.

[Ignitionnet]

true I can make a cup of tea whilst downloading, however if the throttling is over the top (like most isps do) and goes at isdn speeds, I can probably drive down to london for a day trip and come back before its finished :)

streaming over http obviously is time sensitive but rather I meant just loading something like a web article. It is time sensitive in the manner you need to wait for it to load so can read the page, but that is the same as you cannot use what you download until it is downloaded.

The driver I am reffering to in this case is driverpacks for windows and also packages used on customised installation discs (all legal) some can be quite small and some can be a couple hundred meg. I tend to download them only when I need them such as when installing windows on some machines and wanting the most up to date integrated stuff. Granted the last time I did this was probably 2008 but like you said I will wait until I dish out proper judgement.

I guess what you are reffering to is like what I had on legacy VM, browsing on that was noticebly slow yes I wont lie, it was inconveniant but was useable. The real problem then was the lack of ability to use streaming sites, watch F1 online etc. and hideous download speeds. Waiting an extra couple of secs for a page is nothing compared to other problems.

If it was a simple prioritisation system rather than this max 25% of capacity put out, then that would be more reasonable.

None of your above examples though change that downloading is done in isolation and where it's multiple files you're downloading is usually done as a bulk, non-interactive operation while surfing the web depends on responses being received from the previous requests before the next one can be made.

We'll see what impact things have to be honest, neither you nor I know the specific details or what the eventual effects will be of the policies.

[broadbandking]

I am not that fussed really, if you download something in off peak times then full speed a head, All ISP's are going or have start using shaping as the internet is getting bigger and fast speeds means more ISP costs.

[|Kippa|]

For those who torrent you can set your downloads to happen during off peak times when using utorrent with the scheduler. That way you will be able to download at full speeds in off peak times when you aren't around.

[Juo]

ive only had xxl 50mb for 6 weeks and now virgin are going to start traffic shaping it.
robbers :mad:
I feel the same, this sucks :td:

[zer0]

so is http not traffic managed anymore? is there a list of what is and isnt managed? (eg ftp)

would i need a program that measures traffic only from certain applications?

while annoying the total traffic within certain hours was much easier to keep track of

[Sirius]

I am not that fussed really, if you download something in off peak times then full speed a head, All ISP's are going or have start using shaping as the internet is getting bigger and fast speeds means more ISP costs.

And how long before they move the goal posts yet again. Remember they changed the times not long back because the STM changed the way people downloaded ;).

---------- Post added at 19:38 ---------- Previous post was at 19:35 ----------

so is http not traffic managed anymore? is there a list of what is and isnt managed? (eg ftp)

would i need a program that measures traffic only from certain applications?

while annoying the total traffic within certain hours was much easier to keep track of

It looks like any ftp traffic i dont send via my vpn is managed in the evening :(

[zer0]

only getting 400k now on VM usenet:(

[fiferz]

only getting 400k now on VM usenet:(

Getting 4.9MBps just now. I haven't seen any changes yet.

[Ignitionnet]

It looks like any ftp traffic i dont send via my vpn is managed in the evening :(

Could you provide a traceroute to the FTP server - you can remove the destination and any non-VM hops.

[Mick Fisher]

My 20meg was up and down all Thursday afternoon and Friday evening.

When it finally stabilised I ended up with a maximum download of 10meg, determined by DL'ing those 2 demos that TS recommend. Browsing was annoyingly slow, a test torrent ran at .5 to 4.5 meg and a test NNTP download at the same. I switched over to my BE connection and although it maxes out at only 10meg everything returned to normal. The same test Torrent and NNTP DL maxed out the con and browsing was blazingly fast compared to VM.

I might add I was not STM'ed having only DL'ed a few MB.

If this is the result of application throttling then once again VM BB is crippled by restrictions to the point of not being fit for the purpose IMO.

I'm no longer a heavy downloader and am only an occasional torrent user, STM rarely impacted anything I wanted to do however this new regime of restrictions is so severe that £24 for 10MB BE unlimited vs a crippled VM 20meg for £37 is a no brainer.

Unless this is down to a fault with my connection then it looks like, after 15 years, with Cable, VirginMedia have finally driven me to leave. I only have VM BB left having dumped the TV and phone ages ago.

I expect others may follow in droves.

[Ignitionnet]

Your connection was buggered for over a day beforehand and you report browsing was abysmal. I'd suggest that throttling is the last thing you should be pointing the finger at.

[Sirius]

Could you provide a traceroute to the FTP server - you can remove the destination and any non-VM hops.

Dont need to m8, I have found out why and it involved a problem between seat and keyboard. :blush:

All is well now and back to downloading from a private FTP at full wack.

[Rik]

Im going to complain immediately to VM, my Linux BluRay iso is only coming down at 6MB/s (NNTP)

Totally unacceptable ;)