BufferBloat

What a read. Makes absolutely perfect sense. This actually might explain some of the "non-explainable" sporadic problems in networking.

As for the ISPs not knowing about it, I wonder if it's more about them not looking into it so they can claim ignorance.
 
I'll give the ISPs the benefit of the doubt for the MOST part; I'll refer to the "invisibility" of the problem till he had started examining the packet RTT, up AND downstream. But, on the other hand I wouldn't be surprised to find that some ISPs just don't have managers/techs who care, or have the knowledge to handle such investigations. The screenshot with all the "ack dup" lines in it was a shocker.
 
I'll give the ISPs the benefit of the doubt for the MOST part; I'll refer to the "invisibility" of the problem till he had started examining the packet RTT, up AND downstream. But, on the other hand I wouldn't be surprised to find that some ISPs just don't have managers/techs who care, or have the knowledge to handle such investigations. The screenshot with all the "ack dup" lines in it was a shocker.

You can be assured that large ISPs are doing far more network monitoring than what the blog writer has done. I find it hard to believe that something like this would be news to them (or to the big equipment players like Cisco and Motorola). I think because it's a non-issue for most consumers that they really can't be bothered with trying to fix or negate the problem.

It also could help them. The article states that first party VOIP services aren't affected, but third party services are because of the way cable ISPs segment and prioritize their own traffic on their own networks.
 
Back
Top