From: Bit Gossip (bit.gossip@chello.nl)
Date: Thu Jun 21 2007 - 13:55:07 ART
Hi Tarun,
the link is very interesting but I find a contraddiction in it:
- on pag 5-5 it says: 'Fragment_Size = (Max_Allowed_Jitter *
Link_Speed_in_kbps) / 8'
- 'the rate of the PVC CIR is the clocking rate to reference when estimating
fragmentation requirements in a frame-relay environment'
So which one is the rate to use to calculate the fragment size:
link-speed=port access-rate or PVC cir?
Thanks,
Bit.
----- Original Message -----
From: "Tarun Pahuja" <pahujat@gmail.com>
To: "Eric Poulin" <epoulins@yahoo.com>
Cc: <ccielab@groupstudy.com>
Sent: Thursday, June 21, 2007 10:19 AM
Subject: Re: FRTS Fragmentation on a subinterface?
> Eric,
> On all Point-to-Point Protocol (PPP) and Multilink PPP (MLP)
> links,
> TX-ring
> buffer size is automatically configured, and you cannot change these
> default
> buffer values.On Frame Relay links, the TX-ring is for the main interface,
> which all
> subinterfaces also use. The default TX-ring buffer size is 64 packets(Per
> Main T1 interface). You mightneed to change this setting when the
> subinterface is very small or there are many
> subinterfaces.
>
> http://www.ssuet.edu.pk/~amkhan/cisco/Cisco_IP_Telephony_QoS_Design_Guide.pdf
>
> HTH,
> Tarun Pahuja
> CCIE#7707(R&S,Security,Voice,SP,Storage),CCSI
>
>
> On 6/21/07, Eric Poulin <epoulins@yahoo.com> wrote:
>>
>> Hi GS,
>>
>> I have a frame-relay link with two point-to-point subinterfaces, s0/0.501
>> and s0/0.502,
>> and I need to reduce serialization delay for any packets to 10ms on
>> s0/0.501. I know I
>> can accomplish this by using FRTS and making sure the interval is 10ms
>> and
>> also enabling
>> fragmentation to whatever byte size is 10ms, but should I enable it on
>> both
>> subinterfaces?
>>
>> From what I understand, there's a single Tx Ring per interface,
>> considering this, if
>> s0/0.502's interval is left to default (125ms) and is transmitting a 1500
>> byte packet,
>> s0/0.501 would have to wait for that packet to exit the Tx Ring before
>> being able to
>> transmit anything? In this case, s0/0.501 would not meet the requirements
>> of 10ms delay?
>> Is my theory right?
>>
>> Your feedback is greatly appreciated,
>>
>> Thanks,
>> Eric
>>
>>
>>
>> ____________________________________________________________________________________
>> Park yourself in front of a world of choices in alternative vehicles.
>> Visit the Yahoo! Auto Green Center.
>> http://autos.yahoo.com/green_center/
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sun Jul 01 2007 - 17:24:50 ART