dslreports logo
Search similar:


uniqs
7396
pcunite
join:2010-04-10

pcunite to JohnInSJ

Member

to JohnInSJ

Re: Bufferbloat-long read.

said by JohnInSJ
Large buffers provide performance advantages to 99.999% of all home broadband users. The ones saturating their uplinks should know how to fix it.

Incorrect, but we're all here to learn.
pcunite

pcunite to FirebirdTN

Member

to FirebirdTN
Awesome, thank you for posting about this! More people need to know.

Johkal
Cool Cat
MVM
join:2002-11-13
Pennsyltucky

Johkal

MVM

Need to know what?
pcunite
join:2010-04-10

pcunite

Member

said by Johkal:

Need to know what?

Need to know how I greatly increased my Internet service responsiveness by simply preventing my SB6141 cable modem from having its internal buffer filled!
pcunite

pcunite to Johkal

Member

to Johkal
Here is very old document that explains it well:
»www.faqs.org/docs/Linux- ··· WTO.html
ehren8879
join:2013-03-07
Easton, MD

ehren8879 to FirebirdTN

Member

to FirebirdTN
OP, I discovered and corrected the same issue you describe. However the fix was specific to the configuration of our CMTS.

Note: I'm not a Comcast employee nor was this on Comcast's network.

To recap, we could max out the upstream qos of a DOCSIS 2.0 modem without affecting the network from surfing web pages, albeit more slowly than normal. However once we did the same test, with same qos, but on a D3.0 modem network connectivity was pretty much nonexistent. Same as you described.

We have a single Cisco UBR 10K driving our DOCSIS plant. The fix was related to rate-adapt and rate-limit config options we had set on our cable interfaces.

I'm not saying that the buffer bloat algorithm wouldn't improve performance, but it's likely not the cause of the issue you're seeing.

I'd try to run this up the chain at Comcast.