Re: [OT] BigPond You Tube throttling
I'm on optus and for a brief period I tried their 100MB option. I noticed youtube videos would never fully download. It would load ahead a short amount and then stop and just keep pace with the video. I wonder what affect this has on the numbers. David "If we can hit that bullseye, the rest of the dominoes will fall like a house of cards... checkmate!" -Zapp Brannigan, Futurama On 31 October 2012 10:55, Tony Wright wrote: > Hi all, > > > > > > > > I’m interested in seeing how many of you experience YouTube throttling. > > > > It irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any time > I visit YouTube or ABC iView, I get atrocious speeds. I never got this when > I was on TPG. > > > > If you use youtube, there is a youtube page where you can see a comparison > to see how throttling is affecting you: > > http://www.youtube.com/my_speed > > > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > > > > Bring on the NBN I say. > > > > > > > > Regards, > > Tony > > > >
Re: [OT] BigPond You Tube throttling
On Wed, Oct 31, 2012 at 10:55 AM, Tony Wright wrote: > Hi all, > > ** ** > > > > ** ** > > I’m interested in seeing how many of you experience YouTube throttling.*** > * > > ** > Is it deliberate throttling, or what I'd call accidental (caused maybe by inadequate backhaul)? > ** > > It irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any > time I visit YouTube or ABC iView, I get atrocious speeds. I never got this > when I was on TPG. > > ** ** > > If you use youtube, there is a youtube page where you can see a comparison > to see how throttling is affecting you: > > http://www.youtube.com/my_speed > > ** > It doesn't look good, but I am at work. I don't have any expectation that youtube should be good here. > ** > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > I'll look at it on my vdsl2 connection at home, I'm paying for 100/30 there. vdsl seems to work better than adsl, but I think it's the typical installations (fibre to apartment basement, cat 5 to vdsl2 modem in apartment.) that fix the degradation with distance adsl suffers. Cable would be HFC? THat depends on how many users you have running off the local fibre node? I refer you to Graeme Samuel http://www.zdnet.com/fttn-hfc-are-dead-end-nbn-alternatives-samuel-705901/ HFC networks have gotten so bad that Samuel joked that he advises neighbours at the street Christmas party not to subscribe to Telstra's "hopeless" HFC — just so "at least I can continue to get reasonable speeds at 8 p.m." "Fortunately, most people listen to me as the local expert on the subject, and don't go onto Telstra's cable for their broadband," he laughed. "It's a great relief in my street." > > > ** ** > > Bring on the NBN I say. > > ** ** > > > > ** ** > > Regards, > > Tony > > ** ** > > ** ** > -- Meski http://courteous.ly/aAOZcv "Going to Starbucks for coffee is like going to prison for sex. Sure, you'll get it, but it's going to be rough" - Adam Hills
RE: [OT] BigPond You Tube throttling
I have noticed my youtube being slow compared to my internet connectioni think you are right. I am on bigpond liberty as well. I will be looking for a alternative youtube anyay, too many ads are showingagain..getting too powerful for my liking... From: ozdotnet-boun...@ozdotnet.com [mailto:ozdotnet-boun...@ozdotnet.com] On Behalf Of mike smith Sent: Wednesday, 31 October 2012 11:53 AM To: ozDotNet Subject: Re: [OT] BigPond You Tube throttling On Wed, Oct 31, 2012 at 10:55 AM, Tony Wright wrote: Hi all, I’m interested in seeing how many of you experience YouTube throttling. Is it deliberate throttling, or what I'd call accidental (caused maybe by inadequate backhaul)? It irritates me that I have BigPond Cable Ultimate Liberty, experiencing download speeds of between 50Mbps and 100Mbps most of the time, but any time I visit YouTube or ABC iView, I get atrocious speeds. I never got this when I was on TPG. If you use youtube, there is a youtube page where you can see a comparison to see how throttling is affecting you: http://www.youtube.com/my_speed It doesn't look good, but I am at work. I don't have any expectation that youtube should be good here. Basically, normal internet use would see lots of peaks and troughs. When being throttled, the lines flatten out somewhat, and are below the average. My speeds are consistently below Victorian average, and I’m on Telstra’s fastest product. WTF? I'll look at it on my vdsl2 connection at home, I'm paying for 100/30 there. vdsl seems to work better than adsl, but I think it's the typical installations (fibre to apartment basement, cat 5 to vdsl2 modem in apartment.) that fix the degradation with distance adsl suffers. Cable would be HFC? THat depends on how many users you have running off the local fibre node? I refer you to Graeme Samuel http://www.zdnet.com/fttn-hfc-are-dead-end-nbn-alternatives-samuel-705901/ HFC networks have gotten so bad that Samuel joked that he advises neighbours at the street Christmas party not to subscribe to Telstra's "hopeless" HFC — just so "at least I can continue to get reasonable speeds at 8 p.m." "Fortunately, most people listen to me as the local expert on the subject, and don't go onto Telstra's cable for their broadband," he laughed. "It's a great relief in my street." Bring on the NBN I say. Regards, Tony -- Meski <http://courteous.ly/aAOZcv> http://courteous.ly/aAOZcv "Going to Starbucks for coffee is like going to prison for sex. Sure, you'll get it, but it's going to be rough" - Adam Hills
Re: [OT] BigPond You Tube throttling
On Wed, Oct 31, 2012 at 12:18 PM, wrote: > I have noticed my youtube being slow compared to my internet > connectioni think you are right. I am on bigpond liberty as well. > > ** ** > > I will be looking for a alternative youtube anyay, too many ads are > showingagain..getting too powerful for my liking... > > ** > When they get that big, the server/bandwidth costs mandate ads. > ** > > *From:* ozdotnet-boun...@ozdotnet.com [mailto: > ozdotnet-boun...@ozdotnet.com] *On Behalf Of *mike smith > *Sent:* Wednesday, 31 October 2012 11:53 AM > *To:* ozDotNet > *Subject:* Re: [OT] BigPond You Tube throttling > > ** ** > > On Wed, Oct 31, 2012 at 10:55 AM, Tony Wright wrote:* > *** > > Hi all, > > > > > > > > I’m interested in seeing how many of you experience YouTube throttling.*** > * > > ** ** > > Is it deliberate throttling, or what I'd call accidental (caused maybe by > inadequate backhaul)? > > > > > > It irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any > time I visit YouTube or ABC iView, I get atrocious speeds. I never got this > when I was on TPG. > > > > If you use youtube, there is a youtube page where you can see a comparison > to see how throttling is affecting you: > > http://www.youtube.com/my_speed > > > > ** ** > > It doesn't look good, but I am at work. I don't have any expectation that > youtube should be good here. > > ** ** > > > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > > ** ** > > I'll look at it on my vdsl2 connection at home, I'm paying for 100/30 > there. vdsl seems to work better than adsl, but I think it's the typical > installations (fibre to apartment basement, cat 5 to vdsl2 modem in > apartment.) that fix the degradation with distance adsl suffers. Cable > would be HFC? THat depends on how many users you have running off the > local fibre node? I refer you to Graeme Samuel > > ** ** > > > http://www.zdnet.com/fttn-hfc-are-dead-end-nbn-alternatives-samuel-705901/ > > > ** ** > > HFC networks have gotten so bad that Samuel joked that he advises > neighbours at the street Christmas party not to subscribe to Telstra's > "hopeless" HFC — just so "at least I can continue to get reasonable speeds > at 8 p.m." > > "Fortunately, most people listen to me as the local expert on the subject, > and don't go onto Telstra's cable for their broadband," he laughed. "It's a > great relief in my street." > > > > > > Bring on the NBN I say. > > > > > > > > Regards, > > Tony > > > > > > > > > > ** ** > > -- > Meski > > http://courteous.ly/aAOZcv > > > "Going to Starbucks for coffee is like going to prison for sex. Sure, > you'll get it, but it's going to be rough" - Adam Hills > > ** ** > -- Meski http://courteous.ly/aAOZcv "Going to Starbucks for coffee is like going to prison for sex. Sure, you'll get it, but it's going to be rough" - Adam Hills
Re: [OT] BigPond You Tube throttling
Late to the thread but "Cable is contended" is a furphy. On Wed, Oct 31, 2012 at 9:55 AM, Tony Wright wrote: > I’m interested in seeing how many of you experience YouTube throttling. It > irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any > time I visit YouTube or ABC iView, I get atrocious speeds. I never got this > when I was on TPG. > That is what the Internet does. I would think that YouTube would be a small amount of traffic for BP vs BitTorrent etc. More likely explanation: Telstra is notorious for not participating in any sort of Internet Exchange schemes where as other ISPs like TPG do. You would need to look at the path the traffic takes from you to whatever CDN server you're accessing on YouTube and compare that to TPG to understand the topological differences - and that path will change like the weather anyway based on what a variety of parties do in regard to traffic engineering. > ** > > **If you use youtube, there is a youtube page where you can see a > comparison to see how throttling is affecting you: > > ** > > http://www.youtube.com/my_speed > I get 13.37 mbps out of a 45 mbps link here at work. Looking at the graph it fluctuates between 19 and 8 - so proves two tenths of fk all really as the link speed isn't changing like that. Based on that I would probably blame iiNet business for throttling YouTube too. > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > Here's the thing: All of this talk about 'cable being shared' is based on all too common assumptions about the topology and contention ratios of resi-grade broadband products. In an HFC network, yes you do share the C part of that with everyone else in the same run, but as soon as you make it to the exchange, you are also sharing the F part of that with everyone else. Resi-grade contention ratios are typically 80:1 to 250:1. That means, if they have 80 subscribers with 20mbps links, they will have a 20mbps link out of the exchange - or if they're really cheap and nasty they might have a 20mbps link out of the exchange per 250 customers! I don't know what the average backhaul out of an exchange is but I doubt it would be much beyond 100mbps - that stuff costs money. You can get crappy IP Transit for 30 bucks a megabit a month and good stuff for 60 bucks a megabit a month in Oz. So 100mbps uncontended link will be 3000-6000 bucks a month delivered to a DC somewhere central. Yes, this is not what the ISPs charge themselves but it should give you some context to consider when you think about how networks are planned. There simply is not 1-10gbps links into each exchange with bandwidth for everyone all the time. You can buy uncontended business grade DSL services, but you won't get much change out of 500-1000 bucks a month as they have to allow for your bandwidth utilisation all the way back to the core of their network. Even then, it doesn't really make much sense unless you're building a corporate WAN because the whole Internet is contended anyway and you'll lose any sort of traffic prioritisation/special treatment as soon as you leave their network. The only reason the contention ratios work in Oz is that we have tight quotas so you can't really clog the backhaul too much. In the US apparently cable is awful because they have unlimited cable plans - but the ISPs doing that have made a rod for their own back. > > Bring on the NBN I say. > Today you have a very low latency 100mbps HFC link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. On the NBN you will have a very low latency 100mbps fibre link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. Even the last mile in the NBN is contended - just that the numbers are higher. Anyway, I doubt contention at the head-end of your HFC segment is your issue at all, certainly not in relation to YouTube. Even if they shaped YouTube, which I doubt they would be bothered to, I doubt they would be policing it for HFC customers and differently to anyone else. It uses stuff all bandwidth really. David. -- David Connors da...@connors.com | M +61 417 189 363 Download my v-card: https://www.codify.com/cards/davidconnors Follow me on Twitter: https://www.twitter.com/davidconnors Connect with me on LinkedIn: http://au.linkedin.com/in/davidjohnconnors
RE: [OT] BigPond You Tube throttling
That still doesn't explain why my speedtest.net consistently shows speeds of 50Mbps+, yet the youtube speeds are exceptionally poor at the same time. I can even download from other sites and get those downloads transferring at the expected speed (close to what's shown in speedtest or optus broadband test), while that download has very little impact on the download speed of youtube videos. T. From: ozdotnet-boun...@ozdotnet.com [mailto:ozdotnet-boun...@ozdotnet.com] On Behalf Of David Connors Sent: Monday, 5 November 2012 9:20 AM To: ozDotNet Subject: Re: [OT] BigPond You Tube throttling Late to the thread but "Cable is contended" is a furphy. On Wed, Oct 31, 2012 at 9:55 AM, Tony Wright mailto:tonyw...@gmail.com> > wrote: I'm interested in seeing how many of you experience YouTube throttling. It irritates me that I have BigPond Cable Ultimate Liberty, experiencing download speeds of between 50Mbps and 100Mbps most of the time, but any time I visit YouTube or ABC iView, I get atrocious speeds. I never got this when I was on TPG. That is what the Internet does. I would think that YouTube would be a small amount of traffic for BP vs BitTorrent etc. More likely explanation: Telstra is notorious for not participating in any sort of Internet Exchange schemes where as other ISPs like TPG do. You would need to look at the path the traffic takes from you to whatever CDN server you're accessing on YouTube and compare that to TPG to understand the topological differences - and that path will change like the weather anyway based on what a variety of parties do in regard to traffic engineering. If you use youtube, there is a youtube page where you can see a comparison to see how throttling is affecting you: http://www.youtube.com/my_speed I get 13.37 mbps out of a 45 mbps link here at work. Looking at the graph it fluctuates between 19 and 8 - so proves two tenths of fk all really as the link speed isn't changing like that. Based on that I would probably blame iiNet business for throttling YouTube too. Basically, normal internet use would see lots of peaks and troughs. When being throttled, the lines flatten out somewhat, and are below the average. My speeds are consistently below Victorian average, and I'm on Telstra's fastest product. WTF? Here's the thing: All of this talk about 'cable being shared' is based on all too common assumptions about the topology and contention ratios of resi-grade broadband products. In an HFC network, yes you do share the C part of that with everyone else in the same run, but as soon as you make it to the exchange, you are also sharing the F part of that with everyone else. Resi-grade contention ratios are typically 80:1 to 250:1. That means, if they have 80 subscribers with 20mbps links, they will have a 20mbps link out of the exchange - or if they're really cheap and nasty they might have a 20mbps link out of the exchange per 250 customers! I don't know what the average backhaul out of an exchange is but I doubt it would be much beyond 100mbps - that stuff costs money. You can get crappy IP Transit for 30 bucks a megabit a month and good stuff for 60 bucks a megabit a month in Oz. So 100mbps uncontended link will be 3000-6000 bucks a month delivered to a DC somewhere central. Yes, this is not what the ISPs charge themselves but it should give you some context to consider when you think about how networks are planned. There simply is not 1-10gbps links into each exchange with bandwidth for everyone all the time. You can buy uncontended business grade DSL services, but you won't get much change out of 500-1000 bucks a month as they have to allow for your bandwidth utilisation all the way back to the core of their network. Even then, it doesn't really make much sense unless you're building a corporate WAN because the whole Internet is contended anyway and you'll lose any sort of traffic prioritisation/special treatment as soon as you leave their network. The only reason the contention ratios work in Oz is that we have tight quotas so you can't really clog the backhaul too much. In the US apparently cable is awful because they have unlimited cable plans - but the ISPs doing that have made a rod for their own back. Bring on the NBN I say. Today you have a very low latency 100mbps HFC link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. On the NBN you will have a very low latency 100mbps fibre link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. Even the last mile in the NBN is contended - just that the numbers are higher. Anyway, I doubt contention at the head-end of your HFC segment is your issue at all, certainly not in relation to YouTube. Even if they shaped YouTube, which I doubt they would be bot
RE: [OT] BigPond You Tube throttling
Oh, and my reference to the NBN is more about providing a competitor to Telstra, who I am disliking more and more as I do business - nothing to do with the capability of NBN (which is still far superior to Telstra Broadband because of the upload speed.) Oh look, the free modem Telstra provided has hobbled VOIP ports. Conflict of interest? At least I can plug another router in that bypasses it I guess. T. From: ozdotnet-boun...@ozdotnet.com [mailto:ozdotnet-boun...@ozdotnet.com] On Behalf Of David Connors Sent: Monday, 5 November 2012 9:20 AM To: ozDotNet Subject: Re: [OT] BigPond You Tube throttling Late to the thread but "Cable is contended" is a furphy. On Wed, Oct 31, 2012 at 9:55 AM, Tony Wright mailto:tonyw...@gmail.com> > wrote: I'm interested in seeing how many of you experience YouTube throttling. It irritates me that I have BigPond Cable Ultimate Liberty, experiencing download speeds of between 50Mbps and 100Mbps most of the time, but any time I visit YouTube or ABC iView, I get atrocious speeds. I never got this when I was on TPG. That is what the Internet does. I would think that YouTube would be a small amount of traffic for BP vs BitTorrent etc. More likely explanation: Telstra is notorious for not participating in any sort of Internet Exchange schemes where as other ISPs like TPG do. You would need to look at the path the traffic takes from you to whatever CDN server you're accessing on YouTube and compare that to TPG to understand the topological differences - and that path will change like the weather anyway based on what a variety of parties do in regard to traffic engineering. If you use youtube, there is a youtube page where you can see a comparison to see how throttling is affecting you: http://www.youtube.com/my_speed I get 13.37 mbps out of a 45 mbps link here at work. Looking at the graph it fluctuates between 19 and 8 - so proves two tenths of fk all really as the link speed isn't changing like that. Based on that I would probably blame iiNet business for throttling YouTube too. Basically, normal internet use would see lots of peaks and troughs. When being throttled, the lines flatten out somewhat, and are below the average. My speeds are consistently below Victorian average, and I'm on Telstra's fastest product. WTF? Here's the thing: All of this talk about 'cable being shared' is based on all too common assumptions about the topology and contention ratios of resi-grade broadband products. In an HFC network, yes you do share the C part of that with everyone else in the same run, but as soon as you make it to the exchange, you are also sharing the F part of that with everyone else. Resi-grade contention ratios are typically 80:1 to 250:1. That means, if they have 80 subscribers with 20mbps links, they will have a 20mbps link out of the exchange - or if they're really cheap and nasty they might have a 20mbps link out of the exchange per 250 customers! I don't know what the average backhaul out of an exchange is but I doubt it would be much beyond 100mbps - that stuff costs money. You can get crappy IP Transit for 30 bucks a megabit a month and good stuff for 60 bucks a megabit a month in Oz. So 100mbps uncontended link will be 3000-6000 bucks a month delivered to a DC somewhere central. Yes, this is not what the ISPs charge themselves but it should give you some context to consider when you think about how networks are planned. There simply is not 1-10gbps links into each exchange with bandwidth for everyone all the time. You can buy uncontended business grade DSL services, but you won't get much change out of 500-1000 bucks a month as they have to allow for your bandwidth utilisation all the way back to the core of their network. Even then, it doesn't really make much sense unless you're building a corporate WAN because the whole Internet is contended anyway and you'll lose any sort of traffic prioritisation/special treatment as soon as you leave their network. The only reason the contention ratios work in Oz is that we have tight quotas so you can't really clog the backhaul too much. In the US apparently cable is awful because they have unlimited cable plans - but the ISPs doing that have made a rod for their own back. Bring on the NBN I say. Today you have a very low latency 100mbps HFC link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. On the NBN you will have a very low latency 100mbps fibre link going into a contended resi-grade network with variable performance and an 80-250:1 contention ratio. Even the last mile in the NBN is contended - just that the numbers are higher. Anyway, I doubt contention at the head-end of your HFC segment is your issue at all, certainly not in relation to YouTube. Even if they shaped YouTube, which I doubt they would be both
Re: [OT] BigPond You Tube throttling
On Mon, Nov 5, 2012 at 1:36 PM, Tony Wright wrote: > That still doesn’t explain why my speedtest.net consistently shows speeds > of 50Mbps+, yet the youtube speeds are exceptionally poor at the same time. > Yeah it does. Speedtest is really well connected. Most of the sites people use to test from participate in internet exchanges so the path to them is very short and the bandwidth plentiful. Even Telstra who don't have their own speedtest.net nodes you can measure perf from. I can even download from other sites and get those downloads transferring > at the expected speed (close to what’s shown in speedtest or optus > broadband test), while that download has very little impact on the download > speed of youtube videos. > Your issue is most likely to do with Telstra's peering with wherever the YouTube CDN nodes are you use. Like I said, I am on a business grade bonded DSL service here and YouTube reports 8-18mbps. My personal VM in Codify's co-lo reports 6.83 mbps on that YouTube thing even though it was a 1gbps cross connect. Go figure. David. > > T. > > ** ** > > *From:* ozdotnet-boun...@ozdotnet.com [mailto: > ozdotnet-boun...@ozdotnet.com] *On Behalf Of *David Connors > *Sent:* Monday, 5 November 2012 9:20 AM > > *To:* ozDotNet > *Subject:* Re: [OT] BigPond You Tube throttling > > ** ** > > Late to the thread but "Cable is contended" is a furphy. > > On Wed, Oct 31, 2012 at 9:55 AM, Tony Wright wrote:** > ** > > I’m interested in seeing how many of you experience YouTube throttling. It > irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any > time I visit YouTube or ABC iView, I get atrocious speeds. I never got this > when I was on TPG. > > ** ** > > That is what the Internet does. I would think that YouTube would be a > small amount of traffic for BP vs BitTorrent etc. > > ** ** > > More likely explanation: Telstra is notorious for not participating in any > sort of Internet Exchange schemes where as other ISPs like TPG do. You > would need to look at the path the traffic takes from you to whatever CDN > server you're accessing on YouTube and compare that to TPG to understand > the topological differences - and that path will change like the weather > anyway based on what a variety of parties do in regard to traffic > engineering. > > > > If you use youtube, there is a youtube page where you can see a comparison > to see how throttling is affecting you: > > http://www.youtube.com/my_speed > > ** ** > > I get 13.37 mbps out of a 45 mbps link here at work. Looking at the graph > it fluctuates between 19 and 8 - so proves two tenths of fk all really as > the link speed isn't changing like that. Based on that I would probably > blame iiNet business for throttling YouTube too. > > ** ** > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > > > > Here's the thing: All of this talk about 'cable being shared' is based on > all too common assumptions about the topology and contention ratios of > resi-grade broadband products. > > ** ** > > In an HFC network, yes you do share the C part of that with everyone else > in the same run, but as soon as you make it to the exchange, you are also > sharing the F part of that with everyone else. > > ** ** > > Resi-grade contention ratios are typically 80:1 to 250:1. That means, if > they have 80 subscribers with 20mbps links, they will have a 20mbps link > out of the exchange - or if they're really cheap and nasty they might have > a 20mbps link out of the exchange per 250 customers! > > ** ** > > I don't know what the average backhaul out of an exchange is but I doubt > it would be much beyond 100mbps - that stuff costs money. You can get > crappy IP Transit for 30 bucks a megabit a month and good stuff for 60 > bucks a megabit a month in Oz. So 100mbps uncontended link will be > 3000-6000 bucks a month delivered to a DC somewhere central. Yes, this is > not what the ISPs charge themselves but it should give you some context to > consider when you think about how networks are planned. There simply is not > 1-10gbps links into each exchange with bandwidth for everyone all the time. > > > ** ** > > You can buy uncontended business grade DSL services, but you won't get > much change out of 500-1000 bucks a month as they hav
Re: [OT] BigPond You Tube throttling
On Mon, Nov 5, 2012 at 3:29 PM, Tony Wright wrote: > Oh, and my reference to the NBN is more about providing a competitor to > Telstra, who I am disliking more and more as I do business – nothing to do > with the capability of NBN (which is still far superior to Telstra > Broadband because of the upload speed.) Oh look, the free modem Telstra > provided has hobbled VOIP ports. Conflict of interest? At least I can plug > another router in that bypasses it I guess. > Indeed you can. There are a fair few DOCSIS modems out there. Cisco even sell DOCSIS HWICs if you want to go completely nuts on it (and have cash to burn). David. > > > T. > > ** ** > > *From:* ozdotnet-boun...@ozdotnet.com [mailto: > ozdotnet-boun...@ozdotnet.com] *On Behalf Of *David Connors > *Sent:* Monday, 5 November 2012 9:20 AM > *To:* ozDotNet > *Subject:* Re: [OT] BigPond You Tube throttling > > ** ** > > Late to the thread but "Cable is contended" is a furphy. > > On Wed, Oct 31, 2012 at 9:55 AM, Tony Wright wrote:** > ** > > I’m interested in seeing how many of you experience YouTube throttling. It > irritates me that I have BigPond Cable Ultimate Liberty, experiencing > download speeds of between 50Mbps and 100Mbps most of the time, but any > time I visit YouTube or ABC iView, I get atrocious speeds. I never got this > when I was on TPG. > > ** ** > > That is what the Internet does. I would think that YouTube would be a > small amount of traffic for BP vs BitTorrent etc. > > ** ** > > More likely explanation: Telstra is notorious for not participating in any > sort of Internet Exchange schemes where as other ISPs like TPG do. You > would need to look at the path the traffic takes from you to whatever CDN > server you're accessing on YouTube and compare that to TPG to understand > the topological differences - and that path will change like the weather > anyway based on what a variety of parties do in regard to traffic > engineering. > > > > If you use youtube, there is a youtube page where you can see a comparison > to see how throttling is affecting you: > > http://www.youtube.com/my_speed > > ** ** > > I get 13.37 mbps out of a 45 mbps link here at work. Looking at the graph > it fluctuates between 19 and 8 - so proves two tenths of fk all really as > the link speed isn't changing like that. Based on that I would probably > blame iiNet business for throttling YouTube too. > > ** ** > > Basically, normal internet use would see lots of peaks and troughs. When > being throttled, the lines flatten out somewhat, and are below the average. > My speeds are consistently below Victorian average, and I’m on Telstra’s > fastest product. WTF? > > > > Here's the thing: All of this talk about 'cable being shared' is based on > all too common assumptions about the topology and contention ratios of > resi-grade broadband products. > > ** ** > > In an HFC network, yes you do share the C part of that with everyone else > in the same run, but as soon as you make it to the exchange, you are also > sharing the F part of that with everyone else. > > ** ** > > Resi-grade contention ratios are typically 80:1 to 250:1. That means, if > they have 80 subscribers with 20mbps links, they will have a 20mbps link > out of the exchange - or if they're really cheap and nasty they might have > a 20mbps link out of the exchange per 250 customers! > > ** ** > > I don't know what the average backhaul out of an exchange is but I doubt > it would be much beyond 100mbps - that stuff costs money. You can get > crappy IP Transit for 30 bucks a megabit a month and good stuff for 60 > bucks a megabit a month in Oz. So 100mbps uncontended link will be > 3000-6000 bucks a month delivered to a DC somewhere central. Yes, this is > not what the ISPs charge themselves but it should give you some context to > consider when you think about how networks are planned. There simply is not > 1-10gbps links into each exchange with bandwidth for everyone all the time. > > > ** ** > > You can buy uncontended business grade DSL services, but you won't get > much change out of 500-1000 bucks a month as they have to allow for your > bandwidth utilisation all the way back to the core of their network. Even > then, it doesn't really make much sense unless you're building a corporate > WAN because the whole Internet is contended anyway and you'll lose any sort > of traffic prioritisation/special treatment as soon as you leave their > network. > > ** ** > > The only reason the contention ratios work in Oz is th