Netgate Discussion Forum
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Search
    • Register
    • Login

    Calculating the required bandwidth for ACK queues for asymetric link

    Scheduled Pinned Locked Moved Traffic Shaping
    53 Posts 10 Posters 99.8k Views
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • D
      dusan
      last edited by

      @eri--:

      Furthermore, i arrived to a conclusion that with a +-10% tolerance i can simplify it to
      log(X) = 0.8 * log(A) so it will be mostly OK for most of users and remove the dependency on the other side.

      I don't think that it can be simplified that way. X in the original formula depends on on both A and B, while X in the simplified formula depends on A only. Thus the simplified formula is not an approximation of the original (they're completely different and non-comparable instead) and we may not talk about any tolerance. I doubt such a simplification would be helpful.

      1 Reply Last reply Reply Quote 0
      • E
        eri--
        last edited by

        Formally they are very different.
        Practically, with that tolerance the values ressemble pretty much the original for the set of bandwidths of home users.

        1 Reply Last reply Reply Quote 0
        • D
          dusan
          last edited by

          That's an unfounded argument.

          Assume, as an example, A = 1 [mb/s]. Using the simplified formula we would obtain X = 1 [mb/s] regardless of the downlink (B).

          Using the original formula, with several possible downlinks B = 1, 2, 5, 10, 20 [mb/s] we have X = 0.0558, 0.0972, 0.2022, 0.3521, 0.6130 [mb/s], respectively.

          Relative error of the simplified formula against the original is 1692%, 929%, 395%, 184%, 63%, respectively. So, the simplified formula is not 10%-tolerated.

          The ratio of X in the last case (B=20) against X in the first case (B=1) is 0.6130/0.0558 = 1099%. This is big enough to see that no single value of X suits both the cases and, consequently, formulas computing such X can never be tolerated.

          1 Reply Last reply Reply Quote 0
          • N
            Nostradamus
            last edited by

            Hmm… I don't understand a shit, hehe hehe

            I like to setup traffic sharper with max 25/10 mbit for all IP's in this IP Range 192.168.100.100-192.168.100.111 with full max speed to all protocols, like P2P, everything. And give those users in this IP Range 192.168.100.112-192.168.100.125 max 10/5 mbit with low priority of P2P user and other traffic use. But still have 100mbit speed on LAN(local network, i have a server).
            Please don't come with those formulas, because i don't understand:) I need pictures, hehe

            Cheers

            1.2-RELEASE
            built on Fri Mar 7 18:49:15 EST 2008
            Intel Celeron 1100 Mhz
            512MB PC-100
            Segate 20GB HD
            eth0: Intel 82801BA/BAM/CA/CAM 10/100 (WAN)
            eth1: 3Com 3c905C-TX/TX-M 10/100 (LAN)

            1 Reply Last reply Reply Quote 0
            • H
              heiko
              last edited by

              Do you have a calculator?

              1 Reply Last reply Reply Quote 0
              • S
                sullrich
                last edited by

                @Nostradamus:

                Hmm… I don't understand a shit, hehe hehe

                I like to setup traffic sharper with max 25/10 mbit for all IP's in this IP Range 192.168.100.100-192.168.100.111 with full max speed to all protocols, like P2P, everything. And give those users in this IP Range 192.168.100.112-192.168.100.125 max 10/5 mbit with low priority of P2P user and other traffic use. But still have 100mbit speed on LAN(local network, i have a server).
                Please don't come with those formulas, because i don't understand:) I need pictures, hehe

                Cheers

                Please do not hijack this conversation.  If you do not understand then simply ignore!

                1 Reply Last reply Reply Quote 0
                • D
                  dusan
                  last edited by

                  @Nostradamus:

                  I like to setup traffic sharper with max 25/10 mbit for all IP's in this IP Range 192.168.100.100-192.168.100.111 with full max speed to all protocols, like P2P, everything.

                  Let's call it Range1.

                  @Nostradamus:

                  And give those users in this IP Range 192.168.100.112-192.168.100.125 max 10/5 mbit with low priority of P2P user and other traffic use.

                  Let's call it Range2.

                  @Nostradamus:

                  But still have 100mbit speed on LAN(local network, i have a server).

                  Please don't come with those formulas, because i don't understand:) I need pictures, hehe
                  Cheers

                  Well here are some pictures.

                  Link hiearchy:

                  –--Mainlink
                          |
                          +----Range1 (10,25) mb/s
                          |
                          +----Range2 (5,10) mb/s

                  Queue hiearchy:

                  –--qWANroot
                          |
                          +----Range1'qWANroot 10 mb/s
                          |      |
                          |      +----Range1'qWANack
                          |      |
                          |      +---- ...
                          |
                          +----Range2'qWANroot 5 mb/s
                                 |
                                 +----Range2'qWANack
                                 |
                                 +---- ...

                  ----qLANroot
                          |
                          +----Range1'qLANroot 25 mb/s
                          |      |
                          |      +----Range1'qLANack
                          |      |
                          |      +---- ...
                          |
                          +----Range2'qLANroot 10 mb/s
                                 |
                                 +----Range2'qLANack
                                 |
                                 +---- ...

                  For Range2 you don't need to use the formula since the required bandwidths are pre-calculated in the table. The column B/A means how many times is downlink faster than uplink. For Range2 it is 10/5 = 2, so have a look at row 2:

                  -The X/A is the required bandwidth of qWANack, i.e. 9.72%.

                  -The Y/B is the required bandwidth of qLANack, i.e. 3.2%.

                  For Range1, B/A = 25/10 = 2.5 which is not listed in the table. If you don't use the formula then you'll have to guess the required bandwidths from the nearest rows (2 and 3) of the table. Thus qWANack should be something between 9.72% and 13.44%, and qLANack should be something between 3.2% and 2.32%.

                  Hope it helps.

                  1 Reply Last reply Reply Quote 0
                  • N
                    Nostradamus
                    last edited by

                    @dusan.

                    Thank you very very much ;) You are the man

                    Cheers

                    1.2-RELEASE
                    built on Fri Mar 7 18:49:15 EST 2008
                    Intel Celeron 1100 Mhz
                    512MB PC-100
                    Segate 20GB HD
                    eth0: Intel 82801BA/BAM/CA/CAM 10/100 (WAN)
                    eth1: 3Com 3c905C-TX/TX-M 10/100 (LAN)

                    1 Reply Last reply Reply Quote 0
                    • M
                      Master One
                      last edited by

                      Looks more like black magic to me too. Has any of this already been implemented in 1.2-RELEASE? I just run the wizard, and this is what came out:```
                                0    No    830Kb qwanRoot
                                0  No 14552Kb qlanRoot
                                1  Yes 1 %  qwandef
                                1  Yes 1 %  qlandef
                      ACK      7  No 25 %  qwanacks
                      ACK      7  No 25 %  qlanacks
                                7  No 25 %  qVOIPUp
                                7  No 25 %  qVOIPDown
                      RED ECN  1  No 1 %  qP2PUp
                      RED ECN  1  No 1 %  qP2PDown

                      I just changed qwanacks to 65%. No idea, if that's any good yet.
                      1 Reply Last reply Reply Quote 0
                      • E
                        eri--
                        last edited by

                        As Scott said in some posts before ask your question in another thread.

                        The answer to your question is a plain NO.

                        Remember that this is just a recommendation and not suitable to every case.

                        Ermal

                        1 Reply Last reply Reply Quote 0
                        • First post
                          Last post
                        Copyright 2025 Rubicon Communications LLC (Netgate). All rights reserved.