========================================= Minutes of the PSAMP session at IETF 65 Wednesday March 22, 15:10 h - 16:10 h Minutes taken by Ralf Wolter ========================================= Packet Sampling Session Chair: Juergen Quittek 0. Session Summary 1. Agenda bashing, WG Status, PSAMP documents status (Juergen Quittek) 2. PSAMP protocol (Benoit Claise) 3. PSAMP information model (Paul Aitken) 4. PSAMP MIB (Juergen Quittek) 5. Wrap up ---------------- Discussed Internet drafts Packet Sampling (PSAMP) Protocol Specifications http://www.ietf.org/internet-drafts/draft-ietf-psamp-protocol-04.txt Information Model for Packet Sampling Exports http://www.ietf.org/internet-drafts/draft-ietf-psamp-info-04.txt Definitions of Managed Objects for Packet Sampling http://www.ietf.org/internet-drafts/draft-ietf-psamp-mib-05.txt RELATED DRAFTS: IPFIX Protocol Specification http://www.ietf.org/internet-drafts/draft-ietf-ipfix-protocol-19.txt Information Model for IP Flow Information Export http://www.ietf.org/internet-drafts/draft-ietf-ipfix-info-11.txt Managed Objects for IPFIX concentrator http://www.ietf.org/internet-drafts/draft-kobayashi-ipfix-concentrator-mib-01.txt ---------------- 0. Session summary This was most probably the last PSAMP session. The PSAMP WG has two documents (framework and packet selection methods) in state 'publication requested'. The remaining three documents (protocol, info model, MIB) are in a rather mature state. PSAMP protocol specification and PSAMP information model are ready for WGLC. Concerning the PSAMP MIB, a split of the document into an IPFIX-generic part (to be handed over to the IPFIX WG) and a PSAMP-specific part was discussed. A decision on how to proceed will be made when feedback from the IPFIX WG is available. ---------------- 1. Agenda bashing, WG Status, IPFIX documents status (Juergen) The first two PSAMP documents (framework and packet selection methods) were submitted for publication as RFC in July 2005 and are still waiting for AD review. The remaining three are quite mature and discussed at this session. ---------------- 2. PSAMP protocol (Benoit Claise) http://www.ietf.org/internet-drafts/draft-ietf-psamp-protocol-03.txt Benoit already reported on version -04 that has been submitted but was not yet posted. In this version all open issues resolved, there are potentially a few editorial improvements to be applied. The solved issues include selection sequence (#12), selection sequence statittics (#8), accuracy report interpretation (#11), and some miscellaneous (encryption, IANA section, hash based filtering...) issues. Version -04 is ready for WG lastg call. ---------------- 3. PSAMP information model (Paul Aitken) http://www.ietf.org/internet-drafts/draft-ietf-psamp-info-04.txt Paul described changes applied since the last version. Changes include 3 new inserted sections (overview, terminology, overloading information elements), reanming of a few inforamtion elements and solving of all but one open issues. The open issue concerns the ipHeaderPacketSection and mplsLabelStackSection The question is whether or not these IEs should also report payload content if the section length is longer than the header or stack size? Proposal variable length? We must be careful with exporting payload ^ security/legal concerns! This was reinforced by other comments, especially SPs are not allowed to capture payload! The issue will be solved on the mailing list as soon as possible. Then a new version will be produced that will enter WG last call. ---------------- 4. PSAMP MIB (Juergen Quittek) http://www.ietf.org/internet-drafts/draft-ietf-psamp-mib-05.txt Juergen proposed to put the PSMAP MIB on hold until IPFIX has decided on the proposal to combine the PSMAP MIB and concentrator MIB and leave the PSAMP specific part as extensions to the IPFIX MIB. This would introduce another delay to PSAMP caused by IPFIX. The proposal is to remove in the other PSAMP documents normative references to the PSAMP MIB. Then at least the other four PSAMP documents can proceed. Nevil Brownlee: Are people willing to work on the combined MIB? Benoit, Paul, and Thomas volunteered. ---------------- 5. Wrap up (Juergen) This was probably the last psamp session, the WG can be closed once all documents are finalized.