On 02/04/16 13:37, Marko Cupać wrote:
On Thu, 4 Feb 2016 12:45:23 +0100 I guess it should be obvious what I want to achieve: If both hosts request max bandwidth at the same time, throttle them both to 5M. While host2 sits idle, give host1 10M. If host2 requests max bandwidth while host1 downloads at 10M, throttle host1 to 5M and give host2 5M.
It doesn't work that way. They will both be allocated whatever is the 'fair share', within the bounds set by the rules you have specified.
Would getting the different outcome (such as not throttling active host1's active transfer from 10M to 5M, and grant host2 5M when it requests max bandwidth) count as proof for defective mechanism?
You would need to come up with a credible measurement regime. Not forgetting to compensate for whatever the relevant sampling rate for evaluation is (the number escapes me at the moment, but do take a peek at the relevant source files while preparing to measure anything).
-- Peter N. M. Hansteen, member of the first RFC 1149 implementation team http://bsdly.blogspot.com/ http://www.bsdly.net/ http://www.nuug.no/ "Remember to set the evil bit on all malicious network traffic" delilah spamd[29949]: 85.152.224.147: disconnected after 42673 seconds.
