The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2004-Feb> msg00060



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

RE: unequal load sharing when forwarding traffic down MPLS TE tunnels using PBR

  • From: "Eric Osborne" <eosborne@cisco.com>
  • Date: Thu, 26 Feb 2004 12:56:35 -0500
  • Importance: Normal
  • Resent-Date: Thu, 26 Feb 2004 13:32:12 -0500
  • To: "'Palmer, Rob A'" <Rob.Palmer@team.telstra.com>, <mpls-ops@mplsrc.com>
  • X-from-outside-Cisco: [161.44.71.173]
  • X-PMX-Version: 4.1.1.86173



 > -----Original Message-----
 > From: Palmer, Rob A [mailto:Rob.Palmer@team.telstra.com] 
 > Sent: Wednesday, February 25, 2004 7:33 PM
 > To: mpls-ops@mplsrc.com
 > Subject: [MPLS-OPS]: unequal load sharing when forwarding 
 > traffic down MPLS TE tunnels using PBR
 > 
 > 
 > Hi 
 > 
 > I have a question regarding the use of policy based routing 
 > and MPLS TE in a Cisco environment. The idea of using policy 
 > based routing to forward traffic down MPLS TE tunnels is 
 > highly appealing due to the flexibility this offers. However 
 > it is not clear to me whether it also possible to have 
 > unequal load sharing between multiple tunnels between the 
 > same source and destination and use PBR to either directly, 
 > or indirectly, forward the traffic. 
 > 
 > Is it possible to achieve this end in the following fashion?
 > 
 > 1. Create two tunnels (Tunnel0, Tunnel1) between router A 
 > and router B 2. Create two static routes at the headend 
 > router which forward traffic with router B ip address into 
 > the two tunnels eg
 >    ip route routerB_ip_address Tunnel0
 >    ip route routerB_ip_address Tunnel1
 > 3. Use PBR to direct traffic matching the policy to 
 > routerB_ip_address eg route-map thataway
 >   match length 3 50
 >    set ip next-hop routerB_ip_address
 > 
 > 

It sounds like it should work; I'll set it up and try it tomorrow and
let the list know.
But be careful with PBR; using it for a full-fleged architecture, as
opposed to a problem-solving device, can be tricky.



eric

 > thanks in advance for any feedback
 > 
 > Rob
 > 
 > -------
 > The MPLS-OPS Mailing List
 > Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
 > Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
 > 

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml