Question about delay in HFSC

  • Hi,

    I got a -hopefully- simple question regarding the hfsc shaper.

    I read the info there:'s_HFSC_explained

    "queue oBrst bandwidth 0Kb hfsc( realtime (160Kb 6000 94Kb) linkshare (135Kb 6000 84Kb))

    1. The service will always get 160Kb of guaranteed throughput for the first 6 seconds, then drop to 94Kb for any remaining time left
    2. The service will get up to an additional 135Kb of throughput for the first 6 second, then drop to up to 84Kb for any remaining time left "

    In Ermal's example about VOIP in his guide:,11986.0.html
    he uses: "m1 = 1.2Kb d = 30 m2 = 64Kb"

    Now my question is, what's the meaning of an m1 value that is so much lower than m2?
    I understand that I can use m1+d to allow burst traffic, but I can't make sense of it when m1 *d is lower than m2. Doesn't m2 apply always?
    Or doesn't an m2 value by itself, without specifying m1 and d, make any guarantees about how fast those packets sizeof(m1) are processed? If it's like that I understand ;)

    Thank you very much!


    Another question: does m1 and d apply per connection or is this global like m2?

  • 0 < m1 < m2 is useless. To be exact: it does not have any practical usage.

    Pls refer to Example 1.2 in,22885.msg117781.html#msg117781 for an explanation.

  • Thank you very much!

Log in to reply