From: Gregory Gombas (ggombas@gmail.com)
Date: Fri Aug 31 2007 - 11:33:43 ART
Scott,
I agree with eicc. In my testing when I used a very small bc (in that
case 50 bytes) packets over 50 bytes in length were getting dropped.
By configuring fragmentation you can chop up that large packet into
small enough bytes to fit thru the bc.
Regards,
Greg
On 8/30/07, Scott Smith <hioctane@gmail.com> wrote:
> I'm hoping someone can shed some light on why my thinking is wrong on this.
>
> Config example: 2 512k PVCs on a T1 port. Each PVC is shaped via FRTS
> to 512k with no burst and the physical interface can run at T1 speed.
>
> Cisco says "you do not need fragmentation because the port speed is
> T1, irregardless of the shaping."
>
> I know all the in and outs of fragmentation/interleaving,
> serialization delay, etc and how this impacts voice traffic. Where
> I've got a problem is the seeming disregard for the amount of traffic
> that will actually pass through the interface. T1 port or not a single
> PVC cannot send more than 512k since it is shaped to that rate.
> Granted, once the traffic passes through shaper and its time to place
> the bits on the wire it'll be serialized at T1 rate (512k of traffic
> will be serialized at T1 rate) . Since a single PVC can only place
> 512k on the T1 port how is it that we completely ignore this fact when
> deciding to frag or not frag? It seems to me, in this case, the port
> speed is irrelevant because we can only send at 512k... OK, I could go
> on and on but I'll spare you all :-)
>
> So, obviously I'm wrong and or confused... lord knows it isn't the
> first time and will not be the last but I'd really like to know why my
> logic is flawed. Thus far no one who says I'm confused (including
> Cisco) has been able to explain why.
>
> TIA!
>
> --
> Scott
> CCIE #17040 (R&S)
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Sep 01 2007 - 11:32:14 ART