Some QOS doubts

[dinesh_ymca]

Hii guys..

some doubt.well we have a command like

1) policy-map voip-policy

class voip-rtp

priority 1280

class voip-control

bandwidth 64

now whats the difference between using bandwidth and priority command..

i know using priority means i m using LLQ.but how is bandwidth command different from it..both are assigning the respective amount of bandwidth to the class,then where the difference lies..

2) "total amount of bandwidth allocated for all classes included in a policy map must not exceed 75 percent of the available bandwidth on the interface. The other 25 percent is used for control and routing traffic. (To override the 75 percent limitation, use the max-reserved bandwidth command.) "

max-reserved bandwidth commandChanges the maximum configurable bandwidth forRSVP, CBWFQ, LLQ, IP RTP Priority, Frame Relay,IP RTP Priority, and Frame Relay PVC Interface Priority Queueing. The default is 75 percent.

for all other like WFQ is not this 75% limit applicable.

also while seeing show interface command on our 7206 ,4 port G703/G704 120 ohm card i see bandwidth as 1.54 (i.e 75% of 2.048) and 1984 where i have configurewd command

int s2/0

bandwidth 1984.

i m not using the command max-reserved bandwidth ..

Please explain this behaviour.

3)nterface hssi 0/0/0

rate-limit output 15000000 2812500 5625000 conform-action transmit exceed-action drop

here 2812500 5625000 are brust-normal and brust-max.

in some examples i have seen like

rate-limit output 15000000 152812500 155625000 conform-action transmit exceed-action drop

are both of above stament same.means in brust-noraml either we specift the totalbandwidth(152812500 ) or just limit above the total bandwidth like 2812500 .

M i right

thanks

Dinesh

[thisisshanky]

CBWFQ is class based Weighted FQ used to apply queuing based on classes. CBWFQ doesnt provide strict priority queuing for real time traffic, and hence LLQ was introduced, in which strict priority queuing is used for real time traffic. In your case the class voip-rtp matches the voip traffic and strict priority queuing needs to be applied for that traffic. The "priority " command, helps in specifying that voip traffic should be dequeued first, (since it has been assigned with high priority using the "priority" command)

While the bandwidth command, doesnt assign priority queuing to the traffic, but just WFQ. So, both the priority and bandwidth commands, does specify the "Guaranteed bandwidth for the traffic", but one is for priority queueing, while the other is for WFQ.

In your case, voip-rtp traffic will be dequeued before, voip-control traffic.

Hope that is clear!

[dinesh_ymca]

thanks a lot .it helped ..can someone tell abt rest of the two doubts.if i m not clear in my question then do tell me

thanks

Dinesh