[Contents] [Prev] [Next] [Index] [Report an Error]

Configuration on the Ingress Router

You must access the tunnel interface to map the PHB IDs to the EXP bits. The E-series router signals this mapping to all routers on the tunnel. You can establish different PHB-ID–to–EXP mappings for different tunnels.

host1(config)#interface tunnel mpls:example

PHB-ID–to–EXP mapping for the best-effort traffic class:

host1(config-if)#tunnel mpls diff-serv phb-id standard 0x0000 exp-bits 0

PHB-ID–to–EXP mapping for the af1 traffic class:

host1(config-if)#tunnel mpls diff-serv phb-id standard 10 exp-bits 1
host1(config-if)#tunnel mpls diff-serv phb-id standard 12 exp-bits 2
host1(config-if)#tunnel mpls diff-serv phb-id standard 14 exp-bits 3

PHB-ID–to–EXP mapping for the af2 traffic class:

host1(config-if)#tunnel mpls diff-serv phb-id standard 18 exp-bits 4
host1(config-if)#tunnel mpls diff-serv phb-id standard 20 exp-bits 5
host1(config-if)#tunnel mpls diff-serv phb-id standard 22 exp-bits 6

PHB-ID–to–EXP mapping for the ef traffic class:

host1(config-if)#tunnel mpls diff-serv phb-id standard 46 exp-bits 7

Define classifier control lists to classify the incoming packets into classifier groups. Although not shown here, for each CLACL you must define the rules that will select the appropriate incoming packets: be, af1, af2, or ef.

host1(config)#classifier-list be-packets
host1(config)#classifier-list af1-packets
host1(config)#classifier-list af2-packets
host1(config)#classifier-list ef-packets

Define a policy that maps the selected packets into traffic classes. For the assured forwarding classes, this example uses rate limit profiles to set the colors.

host1(config)#policy-list classify-packets
host1(config-policy-list)#traffic-class best-effort classifier-group bf-packets
host1(config-policy-list)#traffic-class ef classifier-group ef-packets
host1(config-policy-list)#traffic-class af1 classifier-group af1-packets
host1(config-policy-list)#traffic-class af2 classifier-group af2-packets
host1(config-policy-list)#rate-limit-profile af1-profile classifier-group af1-packets
host1(config-policy-list)#rate-limit-profile af2-profile classifier-group af2-packets
host1(config)#rate-limit-profile af1-profile
host1(config-rate-limit-profile)#committed-rate 6000000
host1(config-rate-limit-profile)#committed-burst 1000000
host1(config-rate-limit-profile)#peak-rate 8000000
host1(config-rate-limit-profile)#peak-burst 1000000
host1(config)#rate-limit-profile af2-profile
host1(config-rate-limit-profile)#committed-rate 8000000
host1(config-rate-limit-profile)#committed-burst 1500000
host1(config-rate-limit-profile)#peak-rate 12000000
host1(config-rate-limit-profile)#peak-burst 1000000

You attach the policy to the ingress interface of the ingress router. As packets arrive, they are classified with the internal traffic class/color combination and forwarded into the appropriate queues in the fabric. When the packets are sent into the tunnel out of the ingress router, the EXP bits are set according to the router-generated policy (in this example called mpls-exp-setting) that the JUNOSe software automatically attached to the tunnel.


[Contents] [Prev] [Next] [Index] [Report an Error]