Urgent! PBR redistribution with OSPF..Please HELP

james.bastnagel · ‎02-02-2010

We are currently testing a new vendor connection while we have the old one in production. I have set up policy based routing on my core switches at the central site and I am running an OSPF process with my provider for our MPLS network. I have a set group of single hosts at my remote sites that need to come into the central site and be routed to this new vendor connection but the traffic is following the default route that I am advertising from my core--instead of that traffic coming to my core and using PBR, it is going directly to the "old" default gateway that is on the same subnet as the providers CE router and my core switch.

How can I tell the providers router with OSPF to route all traffic from these specific hosts to an alternate default gateway?

I am running 2 4948's in my core layer.

Thank you for you reading.....

James

Giuseppe Larosa · ‎02-02-2010

Hello James,

PBR has to be applied on interfaces that receive traffic to be diverted that is on internal Vlan SVI(s)

int vlan X

ip policy route-map mypbr

! vlan X is internal vlan not that towards providers

! note you may need to deny traffic between private ip addresses

access-list 111 deny ip 10.0.0.0 0.0.0.255 10.0.0.0 0.255.255.255

access-list 111 permit ip 10.0.0.0 0.0.0.255 any

route-map mypbr

match ip address 111

set ip next-hop new-provider-ipaddress

dont't use default option set ip default next-hop or routing table is used first

Hope to help

Giuseppe

nsn-amagruder · ‎02-02-2010

Just noticed that this post has a date of Oct 11, 2005. Hope you got it resolved by now. :-)

Another option besides PBR would be to set up a second HSRP/GLBP group at your remote site, if you use it and just point those host to that gateway. They will use that circuit until it fails. To get the core to use the head end circuit, advertise those host routes in OSFP so you core gets the more specific routes over the new circuit.

Just another option.

Aaron

NonStop Networks, LLC

Message was edited by: nsn-amagruder

Jon Marshall · ‎02-02-2010

nsn-amagruder wrote:

Just noticed that this post has a date of Oct 11, 2005. Hope you got it resolved by now. :-)

Another option besides PBR would be to set up a second HSRP/GLBP group at your remote site, if you use it and just point those host to that gateway. They will use that circuit until it fails. To get the core to use the head end circuit, advertise those host routes in OSFP so you core gets the more specific routes over the new circuit.

Just another option.

Aaron

NonStop Networks, LLC

Message was edited by: nsn-amagruder

Aaron

Just for your info, there are 2 dates in each original post - the one on the right hand side is when it was posted so it was only posted today

Jon

Giuseppe Larosa · ‎02-02-2010

Hello Aaron,

we try to provide an answer sooner or later

Best Regards

Giuseppe

james.bastnagel · ‎02-03-2010

Thank you to everyone who responded. I appreciate your time & helpfulness.

Unfortunately none of this worked in the environment I am working in, but we found a workaround by changing our testing process.