From: Daniel Kutchin (daniel@kutchin.com)
Date: Thu Feb 12 2009 - 15:51:47 ARST
Roger -
Both my mails are intended to be interpreted this way:
!! 1. To set the available BW or the port sending rate !!!!
!!!! use the command "srr-queue bandwidth limit" !!!!
!!!! !!!!
!! 2. ABSOLUTE BW is used to derive shaPed rates; !!!!
!!!! "srr-queue bandwidth limit" is used to derive !!!!
!!!! shaRed rates !!!!
!!!! !!!!
!! 3. The "srr-queue bandwidth limit" command caps !!!!
!!!! the sending rate (whether shaped or shared) !!!!
!!!! !!!!
!! 5. The rates of shared queues are derived from !!!!
!!!! remaining BW ("srr-queue bandwidth limit" minus !!!!
!!!! total shaped rates) and shared weights !!!!
!!!! !!!!
!! 6. Shaped queues are served first !!!!
!!!! !!!!
!! 7. If the total of shaped rate (q1,q2,...) exceeds !!!!
!!!! "srr-queue bandwidth limit" then shaped q1 will !!!!
!!!! be served first, then shaped q2, an so on, !!!!
!!!! until "srr-queue bandwidth limit" is used up !!!!
See this example below from a live lab:
SW1 SW2
+-------------------+ +-----+
R1-e0/0-|-prec1 ->output-q2-| | |
R2-e0/0-|-prec3 ->output-q3-|-f0/12-| |- R2
R2-e0/0-|-prec5 ->output-q1-| | |
+-------------------+ +-----+
Case 1: q1 gets all the shaped rate before any other shaped q
when its shaped rate = srr-queue bandwidth limit
SW1:
interface f0/12
speed 10 <- interface BW
duplex full
srr-queue bandwidth share 1 1 1 1 <- ignored, since shaped wt>0
srr-queue bandwidth shape 5 40 40 40 <- q1 needs 2MB
srr-queue bandwidth limit 20 <- Available = 20% =2MB
Result
q1: 2MB
q2: 0
q3: 0
Case 2: Using the same setting above, I stopped q1 traffic
Result
q1: 0
q2: 250k
q3: 250k
-
Daniel
-----Original Message-----
From: Roger RPF [mailto:rpf@bluemail.ch]
Sent: Mittwoch, 11. Februar 2009 17:01
To: 'Daniel Kutchin'
Subject: AW: need help => srr-queue bandwidth limit
Hi Daniel,
Thanks for that, so this means that to calculate the bandwidth used with the
shaped parameter, I have to use the value of the bandwidth limit as source?
This would explain the calculation of your example below. This is a great
explanation you wrote to me a while ago.
But in that example on the otherhand you say that the bandwidth limit only
affects the shared weight value. Thats the reason why I got confused
because I just read this "old mail" this morning again....
So in that case, the statement that srr-queue bandwidth limit 20 only
affects the shared_BW would be wrong, isn't it?
Here the original post from you:
=======================================
Roger ---
One more thing to note: unlike shared weights, SHAPED weights
are ABSOLUTE weights.
ABSOLUTE means you use shape_wt to directly calculate
Q-BW (= 1/shaped_wt * interface_bw )
Borrowing an example from Petr's blog to demonstrate what
BW will be allocated to each Q[ueue]:
http://blog.internetworkexpert.com/2008/06/26/quick-notes-on-the-3560-egress
-queuing/#more-141
mls qos
!
interface FastEthernet0/13
switchport mode access
load-interval 30
speed 10 <--- interface BW
duplex full <--- I added this one
srr-queue bandwidth share 33 33 33 1
srr-queue bandwidth shape 50 0 0 0
srr-queue bandwidth limit 20 <--- Affects only shared_BW
Q1
shaped wt = 50, Shared wt = 33
==> Shared wt is ignored because shaped wt is non-zero.
==> BW = 1/50 * (10Mbps) (Note that "srr-q band lim 20" is also ignored)
==> BW = 200 kbps
Q2-Q4
Shaped_wt =0
shared_wts are Q2:33,Q3:33 and Q4:1
srr-queue bandwidth limit 20 ==> 20% of 10Mbps ==> 2Mbps
==> remaining output BW to share = 2 - 0.2 = 1.8Mbps
==> Output BW (Q2) = 33/(33+33+1) * 1.8 = 887kbps
==> Output BW (Q3) = 33/(33+33+1) * 1.8 = 887kbps
==> Output BW (Q4) = 1/(33+33+1) * 1.8 = 26kbps
OK, if you want to police and don't want to use MQC, then you could consider
the storm-control command on the other switch (it acts on ingress traffic)
--- Daniel ===========================================================================-----Urspr|ngliche Nachricht----- Von: Daniel Kutchin [mailto:daniel@kutchin.com] Gesendet: Mittwoch, 11. Februar 2009 16:45 An: 'Roger RPF' Betreff: RE: need help => srr-queue bandwidth limit
Roger,
> srr-queue bandwidth limit 10
This will limit the available BW on the port to 10%
So all your QOS calculations on the port will be based on this reduced BW.
- Daniel
-----Original Message----- From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of Roger RPF Sent: Mittwoch, 11. Februar 2009 15:24 To: ccielab@groupstudy.com Subject: need help => srr-queue bandwidth limit
Hi group,
Question regarding the srr-queue bandwidth limit command on the 3560. Does this command affect the shaped and the shared values in total? I mean, I could configure a srr-queue bandwidth limit which is lower than the shaped value like:
int fa1/0 bandwidth 100 srr-queue bandwidth limit 10 ==> stands for 10% which would be 10Mb srr-queue bandwidth shape 5 0 0 0 => inverse absolute weight = BW (= 1/shaped_wt * int_bw) = 1/5*100=20Mb
So how does this work? Or does the bandwidth limit command only limit the bandwidth of the shared queues?
Would be really thankful if someone could clear this
regards
Roger
Blogs and organic groups at http://www.ccie.net
This archive was generated by hypermail 2.1.4 : Sun Mar 01 2009 - 09:44:11 ARST