In my experience, the withdrawals start on the next day and last for about 48-72 hours. The normal dose of Tylenol (Acetaminophen) solves the problems for me.
BitTorrent already tends to prefer closer peers (from among a rather random list), but does so in a gradual and heuristic way (the slot and choke algorithm and like-peer matching -- give it 30 minutes or so and the peer pairs in constant trading relationships are likely to be close to one another).
The IETF recently established the TANA working group to use ISP topology information to help guide peer selection. Plus the Azureus plugin called Ono is a working implementation and research project of another method.
But ultimately, closer is not always better and, where rare pieces are concerned, even healthy. Any recognition of nearby peers will have to be a bias rather than a restriction or these ideas won't work very well.
Ah yes, Richard Bennett and the "thousands of connections" fear mongering --
"In fact, BitTorrent and other peer-to-peer applications open thousands of virtual circuits per application.
This can be observed by network sniffing tools such as Wireshark,..."
Tell him to shut up until he can show us the data with reproducible steps by installing these products on representative Comcast lines using either (both) the recommended or the default settings. I WANT TO SEE THE DATA. Otherwise, he's just making this shit up by grossly misconfiguring his system on purpose, like he has all along.
Robb Topolski... who is tired of this trollish shill and his useful fools
I don't exactly cry "Bullshit" on this, but it isn't what the poster thinks it is
.
When you spoof your router's MAC to a random router, your router reboots and its memory clears up. That keeps things running for a couple of days until your router's memory corrupts and you have to repeat the process. I am 99% sure that your problem is not related to Comcast.
When Comcast bought up large systems to become the largest Cable MSO, it did not buy the Internet. Comcast has no right to change how the Internet works -- not one byte of it.
How the world-wide Internet works is defined by all of us, through our participation and trust in the Internet Society and the Internet Engineering Task Force. To ensure interoperability and access for all, changes must be carefully deliberated and standardized there. The responsibility of operating the Internet in accordance with those standards is entrusted to companies providing access to it. It's not Comcast's job to change how the Internet works nor can it decide who or what gets preference upon it.
I haven't seen anything other than the press reports about something to be circulated around the FCC. I am hopeful that when the details are released that it serves to preserve and protect the Internet from those who would abuse their power and change it.
RFC 4084 does exactly what you're asking -- describes the various levels of Internet access. It is not an Internet Standard, it needs more support and willingness to adopt. Currently, I think it has neither.
Tuesday, Comcast Corporation and Pando Networks announced that they will lead the industry to create a "P2P Bill of Rights and Responsibilities" for users and ISPs. With an FCC hearing on Comcast's anti-peer-to-peer practices scheduled for later this week, this is hardly a surprise. Once again, Comcast makes another sweetheart-sounding deal, but at the wrong time, and with the wrong sweetheart.
It takes a special kind of arrogance for a company that sells Internet Access to team up with another company that sells Content Delivery and together decide what rights and responsibilities that the world's Internet users should have.
As in its earlier "deal" with BitTorrent, Inc., Comcast's announcement tuesday doesn't change any of the facts it faces: in 2006, it assured Congress that network neutrality laws were not necessary, saying it would not "deny, delay, or degrade" its customers in order to deal with traffic congestion. Within a year it was caught secretly doing exactly that! Even after a long string of deceptive and deflective statements and tactics, Comcast continues to degrade their traffic tuesday.
As was the case in the BitTorrent "deal," neither Comcast Corporation nor Pando Networks represents the millions of customers and other members of the Internet community who were impacted when Comcast secretly launched its anti-P2P attack.
Tuesday's announcement came less than 48 hours from the US Federal Communication Committee's public hearing at Stanford University. There, the FCC heard from two panels of experts followed by public testimony on the Comcast incident specifically as well as similar industry practices in general.
And, just like in the BitTorrent deal, we also saw Comcast and Pando Networking executives start to explain why tuesday's "deal" signals that Network Neutrality regulation is not needed in the Broadband Marketplace.
Comcast talking = nothing.
This is a company with a sub-prime credibility rating.
Does flow management work? Linux has a range of RED and BLUE implementions. Hold a contest at your local LUG or LAN Gamer's meets, to see who can set it up the best. Flow management also includes ECN. Have you switched that on yet? There's MTUs and window sizes to consider - default works fine most times, but do you understand those controls and when they should be used?
THIS IS THE FREAKIN' PROBLEM: The solutions that are IETF vetted and improved as STANDARDS do not come in the door to the service providers, waving fat perks and kickbacks as an incentive to buy. Neither CABLE TV COMPANIES nor TELEPHONE COMPANIES invested the internet. In their eyes, the only good solutions are new solutions.
Deep Packet Inspection seems like a solution looking for a problem. But if the problem is that the public generally has too much control over the Internet, then Deep Packet Inspection is the answer.
Somehow, we've made it this far despite decades of explosive growth and bandwidth congestion.
Fairness is not the problem. Fairness is the wedge-issue that CATV-ISPs are trying to use to justify their behavior.
I personally like the rudimentary aspects of the weighted fair queuing proposal -- so let's image that we had it. Would Comcast still have a problem with too many upload bytes from too many homes competing for the upload path back to the CMTS? Yes.
The real problem is that CATV-ISPs are at their upper limits and FIOS is currently superiour. Most CATV nets are DOCSIS 1.1, neighborhoods of 400-500 homes sharing 9-10 Mbps back to the CMTS. Meanwhile, they have to compete with FIOS advertising 15/15, 15-20/5, 20/20, or etc. Mbps. Due to TCP and higher-layer protocol RETURN overhead, CATV ISPs can only offer download speeds if they can reserve 5% in the upload pipe -- so their download is limited by their upload. For example, downloading 8 Mbps from NNTP requires around 200 Kbps. Their upload was 256 Kbps. What happened next: to increase their download speed offering, they pushed out configuration files allowing uploads of 384 Kbps! Cost $0.00 -- no new equipment, no neighborhood splits -- just "let's pretend that we have the bandwidth." After all, customers don't care about upload speed, they just want to download.
It was because the last mile is IP over DOCSIS and DOCSIS doesn't provide for congestion very well. So goes the theory. The explanation fails the sniff test, though, because TCP/IP does have congestion control and the same behaviors that DOCSIS congestion exhibits would trigger the congestion handling behaviors in TCP/IP.
The type of congestion described by Richard Bennett is a pile-on congestion from which there is no return. This just isn't happening in the field.
When did it start working again? What problems were you having before it just started working again?
I would like to tie this into some other changes I have been tracking.
Please reply back by email robb(at)funchords.com (Robb Topolski)
Thanks
--Robb
Prof. Timothy Wu, the man who DID first coin the term "Network Neutrality" testified at the hearing, and he seemed perfectly satisfied that discriminating against users' BitTorrent uploads is a fine example of a Network Neutrality violation.
In your example, the incentive is MONEY gained by charging content providers extra fees for carriage and then giving their traffic preferential treatment.
However, in the Comcast example, the incentive is MONEY saved by eliminating BitTorrent traffic and then putting off the new plant installations installations and additional transit fees that would normally have been paid to handle user demand.
So what's the real difference?
And nobody wants an unmanaged un-Qos'ed internet. But most people think that how the Internet works is the job of the IETF and the Internet Standards... who already have defined ways for applications to identify time-critical, jitter-sensitive packets and have defined what carriers should do about them.
Otherwise, how do you write software for an world-wide internet when half-a-dozen ISPs and transit providers on any given path want to "tune" the higher-level protocols to their own secret views on how the Internet ought to be prioritized?
Bennett has some interesting theories, but none of them convinced us that Comcast's RST forgery would prevent congestion problems where dynamic traffic shaping couldn't.
Comcast's problems with bittorrent originate in the local node , which is a collision domain. Since you can't do dynamic traffic shaping in a collision domain I don't see how dynamic traffic shaping can solve the problem. TCP SYN packets are the problem, and there is no mechanism to drop them where the problem lies.
A wireless 802.11 link is a collision domain, yet they can be shaped. So the idea that transmissions that originate in a collision domain cannot be shaped is flawed.
Furthermore, TCP SYN's that go unanswered (such as would be the case in a collision) are handled by TCP at the localhost using a multiplicative binary backoff method. Connection attempts time out based on either client settings or, as a last resort, protocol settings (whichever is shorter). In other words, if Comcast drops a SYN, the retransmit is delayed. If Comcast drops any TCP packet, the retransmit is delayed. Dropping packets is how gateways of all kinds handle congestion and the TCP/IP response to it is to slow down or, eventually, time out.
It may be DOCSIS 1.1 between the CPE and the CMTS, but it enters and exits as TCP/IP and congestion control is not only possible, it's natural.
Slashdot Mirror
Don't abuse your employees and "friend" them for free.
In my experience, the withdrawals start on the next day and last for about 48-72 hours. The normal dose of Tylenol (Acetaminophen) solves the problems for me.
BitTorrent already tends to prefer closer peers (from among a rather random list), but does so in a gradual and heuristic way (the slot and choke algorithm and like-peer matching -- give it 30 minutes or so and the peer pairs in constant trading relationships are likely to be close to one another).
The IETF recently established the TANA working group to use ISP topology information to help guide peer selection. Plus the Azureus plugin called Ono is a working implementation and research project of another method.
But ultimately, closer is not always better and, where rare pieces are concerned, even healthy. Any recognition of nearby peers will have to be a bias rather than a restriction or these ideas won't work very well.
Robb Topolski
You have not, Brett, because no software makes connections like you're describing.
Bring actual facts or shut up.
I'm done replying to your trolls. You're desperate, and wrong, and being foolish.
Ah yes, Richard Bennett and the "thousands of connections" fear mongering --
"In fact, BitTorrent and other peer-to-peer applications open thousands of virtual circuits per application. This can be observed by network sniffing tools such as Wireshark, ..."
Tell him to shut up until he can show us the data with reproducible steps by installing these products on representative Comcast lines using either (both) the recommended or the default settings. I WANT TO SEE THE DATA. Otherwise, he's just making this shit up by grossly misconfiguring his system on purpose, like he has all along.
Robb Topolski ... who is tired of this trollish shill and his useful fools
As usual, half the story. http://fjallfoss.fcc.gov/prod/ecfs/retrieve.cgi?native_or_pdf=pdf&id_document=6520035161
Those are interesting facts. Maybe I'm not so 99% anymore. What city and state are you in?
I don't exactly cry "Bullshit" on this, but it isn't what the poster thinks it is
.
When you spoof your router's MAC to a random router, your router reboots and its memory clears up. That keeps things running for a couple of days until your router's memory corrupts and you have to repeat the process. I am 99% sure that your problem is not related to Comcast.
--Robb Topolski
-- Robb Topolski
the guy who found Comcast blocking BitTorrent uploads
When Comcast bought up large systems to become the largest Cable MSO, it did not buy the Internet. Comcast has no right to change how the Internet works -- not one byte of it.
How the world-wide Internet works is defined by all of us, through our participation and trust in the Internet Society and the Internet Engineering Task Force. To ensure interoperability and access for all, changes must be carefully deliberated and standardized there. The responsibility of operating the Internet in accordance with those standards is entrusted to companies providing access to it. It's not Comcast's job to change how the Internet works nor can it decide who or what gets preference upon it.
I haven't seen anything other than the press reports about something to be circulated around the FCC. I am hopeful that when the details are released that it serves to preserve and protect the Internet from those who would abuse their power and change it.
--Robb Topolski
Tuesday, Comcast Corporation and Pando Networks announced that they will lead the industry to create a "P2P Bill of Rights and Responsibilities" for users and ISPs. With an FCC hearing on Comcast's anti-peer-to-peer practices scheduled for later this week, this is hardly a surprise. Once again, Comcast makes another sweetheart-sounding deal, but at the wrong time, and with the wrong sweetheart.
It takes a special kind of arrogance for a company that sells Internet Access to team up with another company that sells Content Delivery and together decide what rights and responsibilities that the world's Internet users should have.
As in its earlier "deal" with BitTorrent, Inc., Comcast's announcement tuesday doesn't change any of the facts it faces: in 2006, it assured Congress that network neutrality laws were not necessary, saying it would not "deny, delay, or degrade" its customers in order to deal with traffic congestion. Within a year it was caught secretly doing exactly that! Even after a long string of deceptive and deflective statements and tactics, Comcast continues to degrade their traffic tuesday.
As was the case in the BitTorrent "deal," neither Comcast Corporation nor Pando Networks represents the millions of customers and other members of the Internet community who were impacted when Comcast secretly launched its anti-P2P attack.
Tuesday's announcement came less than 48 hours from the US Federal Communication Committee's public hearing at Stanford University. There, the FCC heard from two panels of experts followed by public testimony on the Comcast incident specifically as well as similar industry practices in general.
And, just like in the BitTorrent deal, we also saw Comcast and Pando Networking executives start to explain why tuesday's "deal" signals that Network Neutrality regulation is not needed in the Broadband Marketplace.
Comcast talking = nothing.
This is a company with a sub-prime credibility rating.
Robb Topolski
Apparent link to the Interview mentioned in the article is: http://www.katebulkley.com/18-19neil_berkett.pdf
Really!! Hahahaha. How ironic that my name follows his!
I can confirm the false-alarm in this case.
--Robb Topolski
PLEASE!!! COME SAY THAT VERY THING IN PERSON
WHAT: FCC En-Banc Public Hearing
on Broadband Practices
WHEN: Thursday, April 17th
TIME: 12:00pm to 7:00pm
Dinkelspiel Auditorium,
Stanford University
471 Lagunita Drive
Palo Alto, CA
SaveTheInternet.com
Does flow management work? Linux has a range of RED and BLUE implementions. Hold a contest at your local LUG or LAN Gamer's meets, to see who can set it up the best. Flow management also includes ECN. Have you switched that on yet? There's MTUs and window sizes to consider - default works fine most times, but do you understand those controls and when they should be used?
THIS IS THE FREAKIN' PROBLEM: The solutions that are IETF vetted and improved as STANDARDS do not come in the door to the service providers, waving fat perks and kickbacks as an incentive to buy. Neither CABLE TV COMPANIES nor TELEPHONE COMPANIES invested the internet. In their eyes, the only good solutions are new solutions.
Deep Packet Inspection seems like a solution looking for a problem. But if the problem is that the public generally has too much control over the Internet, then Deep Packet Inspection is the answer.
Somehow, we've made it this far despite decades of explosive growth and bandwidth congestion.
The Best 10-Minute Explanation of the Network Neutrality Issue
John,
Fairness is not the problem. Fairness is the wedge-issue that CATV-ISPs are trying to use to justify their behavior.
I personally like the rudimentary aspects of the weighted fair queuing proposal -- so let's image that we had it. Would Comcast still have a problem with too many upload bytes from too many homes competing for the upload path back to the CMTS? Yes.
The real problem is that CATV-ISPs are at their upper limits and FIOS is currently superiour. Most CATV nets are DOCSIS 1.1, neighborhoods of 400-500 homes sharing 9-10 Mbps back to the CMTS. Meanwhile, they have to compete with FIOS advertising 15/15, 15-20/5, 20/20, or etc. Mbps. Due to TCP and higher-layer protocol RETURN overhead, CATV ISPs can only offer download speeds if they can reserve 5% in the upload pipe -- so their download is limited by their upload. For example, downloading 8 Mbps from NNTP requires around 200 Kbps. Their upload was 256 Kbps. What happened next: to increase their download speed offering, they pushed out configuration files allowing uploads of 384 Kbps! Cost $0.00 -- no new equipment, no neighborhood splits -- just "let's pretend that we have the bandwidth." After all, customers don't care about upload speed, they just want to download.
Heh.
It was because the last mile is IP over DOCSIS and DOCSIS doesn't provide for congestion very well. So goes the theory. The explanation fails the sniff test, though, because TCP/IP does have congestion control and the same behaviors that DOCSIS congestion exhibits would trigger the congestion handling behaviors in TCP/IP.
The type of congestion described by Richard Bennett is a pile-on congestion from which there is no return. This just isn't happening in the field.
Excellent! Kudos to you!
When did it start working again? What problems were you having before it just started working again? I would like to tie this into some other changes I have been tracking. Please reply back by email robb(at)funchords.com (Robb Topolski) Thanks --Robb
Prof. Timothy Wu, the man who DID first coin the term "Network Neutrality" testified at the hearing, and he seemed perfectly satisfied that discriminating against users' BitTorrent uploads is a fine example of a Network Neutrality violation.
... who already have defined ways for applications to identify time-critical, jitter-sensitive packets and have defined what carriers should do about them.
In your example, the incentive is MONEY gained by charging content providers extra fees for carriage and then giving their traffic preferential treatment.
However, in the Comcast example, the incentive is MONEY saved by eliminating BitTorrent traffic and then putting off the new plant installations installations and additional transit fees that would normally have been paid to handle user demand.
So what's the real difference?
And nobody wants an unmanaged un-Qos'ed internet. But most people think that how the Internet works is the job of the IETF and the Internet Standards
Otherwise, how do you write software for an world-wide internet when half-a-dozen ISPs and transit providers on any given path want to "tune" the higher-level protocols to their own secret views on how the Internet ought to be prioritized?
Good observation! I noticed how every ISP mentioned as an example in the article was a Cable TV provider.
A wireless 802.11 link is a collision domain, yet they can be shaped. So the idea that transmissions that originate in a collision domain cannot be shaped is flawed.
Furthermore, TCP SYN's that go unanswered (such as would be the case in a collision) are handled by TCP at the localhost using a multiplicative binary backoff method. Connection attempts time out based on either client settings or, as a last resort, protocol settings (whichever is shorter). In other words, if Comcast drops a SYN, the retransmit is delayed. If Comcast drops any TCP packet, the retransmit is delayed. Dropping packets is how gateways of all kinds handle congestion and the TCP/IP response to it is to slow down or, eventually, time out.
It may be DOCSIS 1.1 between the CPE and the CMTS, but it enters and exits as TCP/IP and congestion control is not only possible, it's natural.
I did not recognize you! What a coincidence! HAHAHA! That's really amazing!!