ACC Bandwidth max decreases from speed tests but wont increase?
Posted: Sun Apr 23, 2017 7:33 pm
I have made an observation, not necessarily a bug, but curious.
Router: wndr3700v1 with gargoyle 1.8
ACC enabled, QOS on up and down (rules available if desired)
What I have noticed is that the max bandwidth (fair link limit) decreases appropriately when using speed tests during the "busy" part of a day. Observing shows the expected baheviour that when pings exceed the time limit, fair link limit decreases. This continues until the ping time gets below the threshold. this seems appropriate and works very well.
The odd part is, the speed limit stays low, seemingly forever, if only speed tests are run. The fair link limit either remains where it is or goes lower (if ping time exceeds that limit). It seems to never increase, even if ping times are well below the limit for the duration of the test.
Then, Starting a download... things change. As the download runs (longer than a speed test) the fair link limit starts to increase until the ping time starts to also increase. And it stays there, tracking the actual bandwidth capacity by ping time.
So the only oddity is that for speed tests (I tried all of them) the length of run time does not seem long enough to trigger an increase in speed, regardless of link performance.
Is this expected?
Router: wndr3700v1 with gargoyle 1.8
ACC enabled, QOS on up and down (rules available if desired)
What I have noticed is that the max bandwidth (fair link limit) decreases appropriately when using speed tests during the "busy" part of a day. Observing shows the expected baheviour that when pings exceed the time limit, fair link limit decreases. This continues until the ping time gets below the threshold. this seems appropriate and works very well.
The odd part is, the speed limit stays low, seemingly forever, if only speed tests are run. The fair link limit either remains where it is or goes lower (if ping time exceeds that limit). It seems to never increase, even if ping times are well below the limit for the duration of the test.
Then, Starting a download... things change. As the download runs (longer than a speed test) the fair link limit starts to increase until the ping time starts to also increase. And it stays there, tracking the actual bandwidth capacity by ping time.
So the only oddity is that for speed tests (I tried all of them) the length of run time does not seem long enough to trigger an increase in speed, regardless of link performance.
Is this expected?