JN0-664: Service Provider Routing and Switching, Professional(JNCIP-SP)

The problem in this scenario is that R1 and R8 are not receiving each other's routes because of private AS numbers in the AS path. Private AS numbers are not globally unique and are not advertised to external BGP peers. To solve this problem, you need to do the following:

Configure loops on routers in AS 65412 and advertise-peer-as on routers in AS 64498. This allows R5 and R6 to advertise their own AS number (65412) instead of their peer's AS number (64498) when sending updates to R7 and R8.This prevents a loop detection issue that would cause R7 and R8 to reject the routes from R5 and R62.

Configure remove-private on advertisements from AS 64497 toward AS 64498 and from AS 64500 toward AS 64499. This removes any private AS numbers from the AS path before sending updates to external BGP peers.This allows R2 and R3 to receive the routes from R1 and R4, respectively3.

To create a summary route for IS-IS, you need to configure a policy statement that matches the prefixes to be summarized and sets the next-hop to discard. You also need to configure a summary-address statement under the IS-IS protocol hierarchy that references the policy statement. In this case, the policy statement leak-v6 is trying to match the prefix 2001:db9:a:fa00::/61 exactly, but this prefix is not advertised by any router in the network. Therefore, no summary route is created. To fix this, you need to delete the longer keyword from the route-filter term and change the prefix length to /61 exact. This will match any prefix that falls within the /61 range. You also need to delete the export statement under protocols isis, because this will export all routes that match the policy statement to other IS-IS routers, which is not desired for a summary route.

BGP uses several attributes to select the best path for a destination prefix. One of these attributes is origin, which indicates how BGP learned about a route. The origin attribute can have one of three values: IGP, EGP, or Incomplete. IGP means that the route was originated by a network or aggregate statement within BGP or by redistribution from an IGP into BGP. EGP means that the route was learned from an external BGP peer (this value is obsolete since BGP version 4). Incomplete means that the route was learned by some other means, such as redistribution from a static route into BGP. BGP prefers routes with lower origin values, so Incomplete is preferred over EGP, which is preferred over IGP.

The vrf-target export parameter is used to specify one or more BGP extended community attributes that are attached to VPN routes when they are exported from a VRF routing instance to remote PE routers. This parameter allows you to control which VPN routes are accepted by remote PE routers based on their import policies. You can specify more than one vrf-target export value for a VRF routing instance to support advanced route filtering or route leaking scenarios.

To reduce the impact of suboptimal routing in OSPF and OSPFv3 after devices reboot, you can use the overload feature to prevent a router from being used as a transit router for a specified period of time. This allows the router to stabilize its routing table before forwarding traffic for other routers. To enable the overload feature, you need to do the following:

For OSPF, configure the overload statement under [edit protocols ospf] hierarchy level. You can also specify a timeout value in seconds to indicate how long the router should remain in overload state after it boots up. For example, set protocols ospf overload timeout 900 means that the router will be in overload state for 15 minutes after it boots up.

For OSPFv3, configure the overload statement under [edit protocols ospf3] hierarchy level. You can also specify a realm (ipv4-unicast or ipv6-unicast) and a timeout value in seconds to indicate how long the router should remain in overload state after it boots up for each realm. For example, set protocols ospf3 realm ipv4-unicast overload timeout 900 means that the router will be in overload state for 15 minutes after it boots up for IPv4 unicast routing.