OSPF WG IETF 64 Minutes 
            Scribes: Padma Pillay-Esnault and Acee 


   - OSPF WG Dcoument Status (Refer to presentation) 
     Acee Lindem

     Acee: Contact me with info on implementations of: 
       - OSPFv3 TE
       - Multi-area adjacencies

          Encourage all to read:
       - OSPFv3 Graceful Restart

          Need reviewers for:
       - OSPFv3 Update (RFC 2740 Respin)

   - OSPF MANET Design Status (Refer to presentation)
     Tom Henderson

       Design team work and all discussions will be moved to wider audience. 
         Boeing technical report and GTNetS code available for all to 
         to examine. 

     Justin Dean: Different simulation scenarios/parameters will favor 
         different flooding optimizations. More simulation
         study is needed to look at unicast route robustness and scalability 
         with motion.

     Emmanuel Baccelli: Topology reduction shouldn't be a primary goal. 

     Tom Henderson: We did discuss this on the OSPF design list. More 
        discussion of this is included in the Boeing technical report.  

     Russ White: Need to consider reliability in simulations as well as 
        pathological situations. 

     Sue Hares: Simulations do not always match real world results. A case in
        point is OLSR. We don't know enough yet and should go forward with
        both flooding optimizations so that they evaluated. 

     Alvaro Retana - Need to decide on scenarios and metrics for evaluation. 
        We need to test mobility scenarios other than wave-point. 
   
     Tom: Simulation is not the only consideration. Much discussion is 
        included in the technical report. 
  
     Acee: Discusssion will be taken to the OSPF List. Maybe we should go
        forward with both algoritms to allow real world deployments and 
        evaluation. 

     Alvaro: 4 different MANET protocols - be good to decide on one. 
 
     Sue: Need to avoid complexity of algorithms as well. 
 
     Padma: Feels OSPF list is the best place for discussions. We need to 
        base the decisions both on simulations and reality. 

     Tom: Wide scale MANET deployment and testing is very expensive.  

   - MANET Extension of OSPF Using CDS Flooding (Refer to presentation)
      Phil Spanola 

     Fred Baker: What is inherently broadcast about a MANET? Frame relay is a
        prime example of a network that was initially incorrectly modeled in
        OSPF. 

     Tom: No optimizations in MDR that are specific to broadcast. 
        Broadcast capabilities will be exploited if available. 
  
     Alvaro: MDRs favor adjacency reduction. However, more neighbors requires
        more time for MDR selection during which time flooding will not be 
        done. In contrast, MPRs have built-in redundancy during transients. 

     Tom: Both algorithms have comparable costs with respect to selection.
          
   - High Mobility Scenarios with MPRs and MDRs (Refer to presentations)
      Stan Ratliff

     Tom: What is the timetable for the release of the code and draft update?

          Random wavepoint is normally the worst case and most agressive 
          scenario. 

          What is the evolution of adjacencies as smart peering is used over
          time? 

     Acee: Draft and code should be available within week or so after IETF. 

     Russ: Mobility has two components - range and area. What is the spareness
           of the topology? 
   
   - OSPF Link Local Signalling as Standards Track (Refer to presentation)
      Acee Lindem