cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
913
Views
5
Helpful
6
Replies

A9K-2X100GE-SE service-policy limit

MilanHSBS
Level 1
Level 1
we have limit only 1000-1100 interfaces with service-policy and than we get warning message

prm_ezhal' detected the 'warning' condition 'Not enough TM resources to satisfyrequest')

 
I think this card should manage much more interfaces. 
 
For example we using also card A9K-24X10GE-SE which can manage around 2600 interfaces.  Processor cards we us A9K-RSP440-SE
 
Do you know where can be problem ? version of IOS XR is 6.5.3
 
Thank you
1 Accepted Solution

Accepted Solutions

Hi,
There is no way to increase the limit, you will need to remove older policies to make room for new ones.
The newer line cards 8x100G or 4x100 do not have this limitation, you can apply around 24k on the lower scale card per npu for v4 per direction instead of around 1100 for the entire card

Thank you

View solution in original post

6 Replies 6

tkarnani
Cisco Employee
Cisco Employee

we will need to check the total resources used on this card

 

show qoshal resource summary location<>

 

this card is limited on the # of 2 level policies that can be configured, the reason for this is that the NPU are shared on this card. 2 for ingress, 2 for egress

 

therefore the policy resources are shared amongst all the NPUs

 

the difference on the 24x10G card it has more NPU's to spread these policies over so you can have more of them

 

hope this helps

 

THanks

 

Hi Tkarnani,

 

thank you for your reply and explain.

 

Attached is export of the command #show qoshal resource summary location<>

 

Maybe you can check the total resources used 

 

We need to know if card is OK or we do some wrong configuration on it. 

 

Thanks

Milan

the line card is fine, its just that its run out of resources. you can see here its full

Policy Instances: Ingress 131 Egress 1000  Total: 1131
   Entities: (L4 level: Queues)
     Level        Chunk 0           Chunk 1           Chunk 2           Chunk 3          
     L4      4017( 4017/ 4017)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L3(8Q)  1004( 1004/ 1004)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L3(16Q)    0(    0/    0)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L2         4(    4/    4)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L1         4(    4/    4)    0(    0/    0)    0(    0/    0)    0(    0/    0)
   Groups:
     Level        Chunk 0           Chunk 1           Chunk 2           Chunk 3          
     L4      1004( 1004/ 1004)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L3(8Q)   254(  254/  254)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L3(16Q)    0(    0/    0)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L2         4(    4/    4)    0(    0/    0)    0(    0/    0)    0(    0/    0)
     L1         4(    4/    4)    0(    0/    0)    0(    0/    0)    0(    0/    0)
   Policers: Internal 0(0) Regular 0(0)  Parent 0(0) Child 0(0) Total 0(0)

Thank you for explain Tkarnani,

 

is there solution to increase that limit ?

we need to buy next card or higher processor card or do any other configuration on service-policy ?

 

thanks for help

Milan

Hi,
There is no way to increase the limit, you will need to remove older policies to make room for new ones.
The newer line cards 8x100G or 4x100 do not have this limitation, you can apply around 24k on the lower scale card per npu for v4 per direction instead of around 1100 for the entire card

Thank you

Thank you Tkarnani for explain