tertech
join:2008-04-12
Ottawa, ON
| reply to CanadianISP
Re: Is your VoIP or VPN affected by Bell's throttling practices?
As much as I dislike the idea of anyone purposely interfering with my ability to do my job, I could not in good conscience make such an accusation without some compelling evidence. Isn't that just the thing we are accusing them of doing with their statement of 5% of users taking xx% of bandwidth?
Is there a software utility like 'Ethereal' that could determine if a packet was purposely dropped or delayed? I think that would be difficult to prove, given the complexity of the network.
BTW... I got a call this evening from Bell trying to get me to switch to Sympatico. I responded like he was trying to sell me leprosy. The guy was taken aback; he didn't know what to say. |
|
R0CKY
TSI Rocky
Premium,VIP
join:2005-05-19
Chatham, ON | I agree... be sure it's affecting you. False claims are going to undermine what Marc said at the start in my mind.
--
TSI Rocky - TekSavvy Solutions Inc. |
|
CanerisErik
Premium
join:2007-10-03
Toronto, ON
| reply to tertech
said by tertech  :Is there a software utility like 'Ethereal' that could determine if a packet was purposely dropped or delayed? Let's consider two possible ways in which the throttling can be done.
If they're sending RSTs, you'd see "more" (very relative and loose) RSTs than during non-throttled times when doing the same thing, but this isn't a guarantee because you can't prove it's not something else causing it.
If they're dropping or delaying ACKs or other data, you'd have "more" retransmits, but again, you can't differentiate that easily from a congested router for example.
I can't think of a way to determine for certain using a packet sniffer if they're throttling. |
|
Name
join:2008-03-28
| It's a bit hard to do TCP RST forgery on UDP or ESP streams. 
Bell is most likely delaying packets in transit.
--
Coridon Henshaw -=- »www.talisiorder.ca |
|
CanerisErik
Premium
join:2007-10-03
Toronto, ON
| said by Name :It's a bit hard to do TCP RST forgery on UDP or ESP streams. Bell is most likely delaying packets in transit. Yes, it is a bit hard
But keep in mind that the majority of P2P is TCP-based, not UDP-based and supposedly that's their focus.
Furthermore, their largely irrelevant response to the CAIP had a funny appendix there teaching everyone "how TCP works" and "how the Internet works", when the issue has zero to do with the Internet and with TCP.
So, it sounds like they're focused on TCP. Perhaps dropping ACKs. |
|
