Internet Engineering Task Force PIM Working Group The PIM Working group met Monday, December 7 in Orlando. The room size was sufficient (not too big, not too small) and the session was multicast on the MBONE. Brian Haberman took a few minutes to discuss the status of his PIM for IPv6 draft. It was decided that the draft should not be incorporated into the standard PIM spec at this time in order not to slow down the standardization process. However, the draft should become an official working group document. Puneet Sharma discussed his "Scalable Timers for PIM" paper. The need for adjusting timers based on the amount of control traffic was well recognized. Puneet's paper recommends a cap on the amount of bandwidth used for control traffic and proposes to lengthen the transmit timers at the sender and estimate the refresh interval at the receiver in order to not exceed the bandwidth cap. The timers are divided into classes and a certain fraction of the bandwidth is guaranteed to each class to avoid starvation. Liming Wei presented a new version of the Authentication draft which merged ideas from previous discussions. There was some reservation about a shared private key in the Differentiated capabilities method and a counter proposal was made to allow each candidate bootstrap router to have its own private key and the public keys of each of the other candidate bootstrap routers. Also, the new draft specifies the Security Parameter Index to be configured along with the keys since there were objections to the previous draft using a reserved SPI. A last call was started on the document to submit it as a proposed standard. Please review this document as soon as possible. Liming then discussed changes to the dense mode spec that were made recently. These were mostly wording changes for clarity. Also, Liming mentioned that he and Bill Fenner had worked out some of the issues brought up by Bill on the mailing list in a private discussion and we can look forward to those changes making it into a new version of the spec soon. The assert vs. prune on p2p links were briefly mentioned and Liming recommended that staying with prunes since the functionality was not broken. Dirk Ooms discussed his draft on "MPLS with PIM-SM". This draft advocates the use of data traffic to trigger the setup of Label Switched Paths instead of setting up the LSPs before hand based on the topology. It also points out some difficult issues with the co-existence of (*,G) state and (S,G) state. It was decided that this work should continue in the MPLS working group.