Question about the BGP route-map cmd 'set metric-type internal'?

[alfred_s_jones]

The documentation (at the URL below) for the BGP route-map action cmd 'set metric-type <blah>' says this command sets the MED value of the matching BGP update to the value of the IGP metric to the "next hop".

http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/fiprrp_r/bgp_r/1rfbgp2.htm#xtocid23

I can see how this works at the edges of a transit AS where the NEXT_HOP values in incoming updates are reachable (if properly configured of course!) via an IGP.

But how does this command work (if its used) in the AS that originates the prefix, i.e. the firs AS that is listed in the AS-PATH? Assume Ra in AS 100 is the originating AS for prefix 192.168.50/24. Assume further that this prefix is redistributed into BGP from a (local) static route. If I'm executing 'set metric-type internal' on all prefixes advertised via eBGP then what is the MED value set to for the update carrying 192.168.50/24? There is no IGP metric associated with the next-hop for this prefix is there?

[DALE FRANCIS]

Your paragraph seems a tad confusing, MED is advertised to a neigbour, neigbuors have to explicitly configured to accept to send MED.

If iBGP derived route then the MED will take the value of the IGP, and advertise that out to iBGP peers. As for the next hop it HAS to be known via an IGP, the prefix's associated with that NEXT_HOP will or can have a MED, Weight etc value to determine the best path to use.

MED is a transitive atttribute and generally exchanged between AS's

May i ask what it is you are trying to achieve.. load balancing..???