Can't seem to dodge throttling at all

[Bob.Dole 0]

Hi all. Just signed up yesterday for 3 months after doing a bit of research and deciding that AirVPN had the principles that I care about and appear to put them into practice with their policy. So far I am very pleased with the service. Much faster than i was expecting when I decided to start looking for a VPN service.

Ok so I am with the UK ISP TalkTalk. Until a while ago I was getting good P2P speeds (close to 2mbps at off peak, few hundred kbps at peak) but then that dropped to at best a few hundred kbps at off peak and pretty much zero (<10kbps) during peak. I asked for an explanation of this from them and they offered nothing in the way of thoughts about it, their customer service really is utter shit. I am pretty sure they just have a policy of "turning down" the speeds of heavy P2P users, even though their policy states explicitly that although they shape traffic they won't punish people based on their total bandwidth usage.

So anyway, joined AirVPN to try and get round that. I have so far tried several different servers, UDP and TCP and each different port option, and my P2P speeds are still the same through VPN as they are without it. Is there anything further that I can try? Can you guys give me any clue (I'm not very techy) about how they can throttle P2P when afaik they can't distinguish the P2P traffic from non-P2P? Am I stuck with these shitty speeds forever? (shitty P2P speeds, not shitty speeds from you guys... the speeds I get with speed tests through AirVPN are brilliant, over 10mbps often, which is what is puzzling me about the throttling)

Thanks for any help and hello from a member of the UK Pirate Party

[Staff 9972]

Hello!

Welcome aboard!

Assuming that your p2p client is effectively tunneled (please make sure it is) the only way your ISP can guess p2p traffic is analyzing your traffic pattern trying to guess from it whether it's p2p or not. The p2p traffic pattern on a device that uses only a p2p client are typical. This system is quite ineffective and normally was used many years ago only by desperate ISPs which had performed such a massive overselling that they didn't know anymore how to handle congestion (we don't know how far in overselling TalkTalk went).

We have studied extensively this system about 10 years ago with a couple of providers, when AirVPN was not even born. The system can be easily fooled by saturating the bandwidth at the beginning of your p2p-ing. Once your p2p client reaches a fairly high bw usage, you should not need anymore to saturate your bandwidth.

Kind regards

[Bob.Dole 0]

Thanks for the very quick response

I'm a newbie to all of this, how would I check if my P2P client is tunnelling correctly? And how would I go about saturating bandwidth when I start P2P client? Do you mean for example open up a bunch of youtube videos and have them loading whilst P2P gets up a bit of speed? (I'll try that just now, just asking in case that's not what you mean).

[Staff 9972]

Thanks for the very quick response

I'm a newbie to all of this, how would I check if my P2P client is tunnelling correctly? And how would I go about saturating bandwidth when I start P2P client? Do you mean for example open up a bunch of youtube videos and have them loading whilst P2P gets up a bit of speed? (I'll try that just now, just asking in case that's not what you mean).

Hello!

Please perform this test:

http://checkmytorrentip.com/

If successful (i.e. you can't see your real IP address), it's easy to saturate your download bandwidth with an http or ftp download of very large file.

ftp will allow you to set bw limits, so that you can progressively share more bandwidth to the p2p client. However, it is not very likely that TalkTalk is applying this desperate and old method (it is dangerous because it is subject to so many false positives that customers could get throttled regardless of what they do - actually we can see that TalkTalk has the highest rate per customers of complaints to Ofcom in the UK for packet shaping reasons...), so there might be different reasons for such p2p bad performance.

Kind regards

[Bob.Dole 0]

Thanks

Client was showing the correct IP (server in Netherlands), tried flooding bandwidth and no luck. Damn TalkTalk, I am dumping them as soon as my contract runs out. Yes I had also seen that they were rated lowest for customer service by OFCOM, unfortunately I never seen this before I signed up

Oh well. No big deal. It's unfortunate that I haven't managed to evade their throttling, but I wanted to start using a VPN anyway just for the sake of privacy.

[Staff 9972]

It's unfortunate that I haven't managed to evade their throttling, but I wanted to start using a VPN anyway just for the sake of privacy.

Hello!

That's very strange. Probably we're missing something here. Unfortunately none of us is in the UK. Is there any TalkTalk customer with an average/high technical knowledge here who would like to perform tests?

Kind regards

[Bob.Dole 0]

I've messaged a person on the TalkTalk forums who seems to know a lot about the technical side of things. Maybe he can shed some light on it.

Ever weirder, if I do a speed test whilst P2P is running I still get a result of 5-10mbps and the P2P traffic is still staying throttled while the speed test runs. I don't know how they're doing it but they definitely seem to be able to differentiate between http traffic and torrent traffic.

I'm still not complaining at all btw, I'm *very* happy with the service from AirVPN. It's funny to me that for such a cheap price you guys are able to offer unshaped, uncapped, unmonitored internet access to someone in a different country... but my ISP (who charge me over 9 times the price of AirVPN) can't give me anywhere near the same service. Am I right in thinking that VPN providers are basically ISP's themselves?

[Staff 9972]

I've messaged a person on the TalkTalk forums who seems to know a lot about the technical side of things. Maybe he can shed some light on it.

Ever weirder, if I do a speed test whilst P2P is running I still get a result of 5-10mbps and the P2P traffic is still staying throttled while the speed test runs. I don't know how they're doing it but they definitely seem to be able to differentiate between http traffic and torrent traffic.

Hello!

This is a proof that you are performing p2p (or possibly everything) outside the tunnel OR more simply that your client is mis-configured. Underlying packets header, payload and size are physically (mathematically) not discernible by your ISP, so it can only either shape everything or nothing on your OpenVPN throughput.

Please send us your OpenVPN client logs.

Kind regards

[Bob.Dole 0]

Thanks, that's good if it is something that is misconfigured... means maybe I can fix it

I can't locate the OVPN client logs. I checked the folder it should be in (using Linux) and it's not there, also searched computer and turned up nothing. I'll keep trying to figure out where it is.

I'm definitely connected as far as I can tell. This site says I am connected, the test that you told me to do with my torrent client says it is connected, my browser homepage is thepiratebay.se and without a vpn that site won't load cos it's censored (that's why I made it my homepage, as a quick way of telling if I am connected before I start browsing).

I tried a different torrent client with no luck as well.

[Staff 9972]

@Bob.Dole

Hello!

Please launch OpenVPN with the --log-append directive followed by path and name of the file you wish to write as log. Alternatively, just launch OpenVPN, copy the shell output and paste it here.

Kind regards

[anonymous1970 0]

I am having the same issues as Bob.Dole except that I am on Centurylink in the U.S.

I tried checkyourtorrentip.com just like you suggested for Bob.Dole and the ip addressed shown was from Air.

These slow speeds are becoming increasingly frustrating. Any suggestions you could provide would be greatly appreciated.

Many thanks.

[Staff 9972]

Hello!

Please try to limit your upload speed. We don't shape traffic at all, so it might be an ADSL problem (if you're on an asymmetric line), since a too high upload bandwidth will "strangle" your download bandwidth.

PeerBlock/PeerGuardian blocking lists have been reported to cause a lot of troubles in our network (because some of those lists block a huge amount of perfectly legitimate IP addresses, including some of our servers IP addresses). If you're running this type of programs, try to disable them.

Finally, if you're using Windows and the Windows firewall, please check that it does not filter your torrent client on ANY network. Even though the torrent client is authorized on some network, it might be filtered in other ones (when you enter the Virtual Private Network, you're in another network).

Kind regards

[Bob.Dole 0]

My torrent client has it's upload speed limited to 35kbps, that was all my ISP allowed it to do anyway, and having it unlimited messed my http trafic up (probably exactly what you are referring to).

I was getting an error when trying to use the --log-append function as suggested earlier, probably something to do with how I am writing the command, I tried several variations with no luck

Here is the output I get when I fire up the VPN. To my untrained eye, everything looks fine... but I have no idea lol.

Mon Nov 5 21:35:40 2012 OpenVPN 2.2.1 x86_64-linux-gnu [sSL] [LZO2] [EPOLL] [PKCS11] [eurephia] [MH] [PF_INET6] [iPv6 payload 20110424-2 (2.2RC2)] built on Mar 30 2012

Mon Nov 5 21:35:40 2012 NOTE: OpenVPN 2.1 requires '--script-security 2' or higher to call user-defined scripts or executables

Mon Nov 5 21:35:40 2012 WARNING: file 'user.key' is group or others accessible

Mon Nov 5 21:35:40 2012 LZO compression initialized

Mon Nov 5 21:35:40 2012 Control Channel MTU parms [ L:1560 D:140 EF:40 EB:0 ET:0 EL:0 ]

Mon Nov 5 21:35:40 2012 Socket Buffers: R=[87380->131072] S=[16384->131072]

Mon Nov 5 21:35:40 2012 Data Channel MTU parms [ L:1560 D:1450 EF:60 EB:135 ET:0 EL:0 AF:3/1 ]

Mon Nov 5 21:35:40 2012 Local Options hash (VER=V4): '958c5492'

Mon Nov 5 21:35:40 2012 Expected Remote Options hash (VER=V4): '79ef4284'

Mon Nov 5 21:35:40 2012 Attempting to establish TCP connection with [AF_INET]95.211.191.33:443 [nonblock]

Mon Nov 5 21:35:41 2012 TCP connection established with [AF_INET]95.211.191.33:443

Mon Nov 5 21:35:41 2012 TCPv4_CLIENT link local: [undef]

Mon Nov 5 21:35:41 2012 TCPv4_CLIENT link remote: [AF_INET]95.211.191.33:443

Mon Nov 5 21:35:41 2012 TLS: Initial packet from [AF_INET]95.211.191.33:443, sid=04c2adef 7530afe7

Mon Nov 5 21:35:42 2012 VERIFY OK: depth=1, /C=IT/ST=IT/L=Perugia/O=airvpn.org/CN=airvpn.org_CA/emailAddress=info@airvpn.org

Mon Nov 5 21:35:42 2012 VERIFY OK: nsCertType=SERVER

Mon Nov 5 21:35:42 2012 VERIFY OK: depth=0, /C=IT/ST=IT/L=Perugia/O=airvpn.org/CN=server/emailAddress=info@airvpn.org

Mon Nov 5 21:35:44 2012 Data Channel Encrypt: Cipher 'AES-256-CBC' initialized with 256 bit key

Mon Nov 5 21:35:44 2012 Data Channel Encrypt: Using 160 bit message hash 'SHA1' for HMAC authentication

Mon Nov 5 21:35:44 2012 Data Channel Decrypt: Cipher 'AES-256-CBC' initialized with 256 bit key

Mon Nov 5 21:35:44 2012 Data Channel Decrypt: Using 160 bit message hash 'SHA1' for HMAC authentication

Mon Nov 5 21:35:44 2012 Control Channel: TLSv1, cipher TLSv1/SSLv3 DHE-RSA-AES256-SHA, 2048 bit RSA

Mon Nov 5 21:35:44 2012 [server] Peer Connection Initiated with [AF_INET]95.211.191.33:443

Mon Nov 5 21:35:46 2012 SENT CONTROL [server]: 'PUSH_REQUEST' (status=1)

Mon Nov 5 21:35:46 2012 PUSH: Received control message: 'PUSH_REPLY,redirect-gateway def1,dhcp-option DNS 10.5.0.1,comp-lzo no,route 10.5.0.1,topology net30,ping 10,ping-restart 120,ifconfig 10.5.0.6 10.5.0.5'

Mon Nov 5 21:35:46 2012 OPTIONS IMPORT: timers and/or timeouts modified

Mon Nov 5 21:35:46 2012 OPTIONS IMPORT: LZO parms modified

Mon Nov 5 21:35:46 2012 OPTIONS IMPORT: --ifconfig/up options modified

Mon Nov 5 21:35:46 2012 OPTIONS IMPORT: route options modified

Mon Nov 5 21:35:46 2012 OPTIONS IMPORT: --ip-win32 and/or --dhcp-option options modified

Mon Nov 5 21:35:46 2012 ROUTE default_gateway=192.168.1.1

Mon Nov 5 21:35:46 2012 TUN/TAP device tun0 opened

Mon Nov 5 21:35:46 2012 TUN/TAP TX queue length set to 100

Mon Nov 5 21:35:46 2012 do_ifconfig, tt->ipv6=0, tt->did_ifconfig_ipv6_setup=0

Mon Nov 5 21:35:46 2012 /sbin/ifconfig tun0 10.5.0.6 pointopoint 10.5.0.5 mtu 1500

Mon Nov 5 21:35:46 2012 /sbin/route add -net 95.211.191.33 netmask 255.255.255.255 gw 192.168.1.1

Mon Nov 5 21:35:46 2012 /sbin/route add -net 0.0.0.0 netmask 128.0.0.0 gw 10.5.0.5

Mon Nov 5 21:35:46 2012 /sbin/route add -net 128.0.0.0 netmask 128.0.0.0 gw 10.5.0.5

Mon Nov 5 21:35:46 2012 /sbin/route add -net 10.5.0.1 netmask 255.255.255.255 gw 10.5.0.5

Mon Nov 5 21:35:46 2012 Initialization Sequence Completed

[Bob.Dole 0]

I am a little more concerned about this than I was before. Initially I was only trying to get it resolved to increase my P2P speeds, but as admin pointed out there should be no way of them distinguishing between traffic on a properly tunnelled connection... and they are definitely managing to tell the difference between my http and P2P traffic, even though everything is telling me that it is all going through the vpn. So this may be a privacy concern as well as just an "I want faster P2P" concern

[Bob.Dole 0]

Just giving this a friendly bump since nearly 5 days have passed