Bulk Transport WG Call Friday, May 13, 2005 Attendees: Larry Dunn Susan Evett (scribe) Yunhong Gu Injong Rhee Chester Ruszczyk Steve Senger Stanislav Shalunov Lisong Xu Call started at 12:00 p.m. EDT. Agenda Review of Previous Action Items a. All (except Stas): Review the draft API Steven sent out (v.6.) and get comments to Steven. [Comments received from Jeff Boote; Chet will review beginning of next week and get comments to Steven by 5/20.] b. Injong to track the COMNET call for papers and keep the group posted when a date has been published. [Ongoing] c. Stas to further review UDT Protocol as a starting point -- Steven will also consider the protocol with reference to his application/API needs. [Ongoing; Stas hasn't had a chance to look at it yet and Gu is on vacation; concern whether we can implement the API on top of UDT. Steven hasn't had a chance to look at the most recent version; he has concerns about partial reliability and the ability to respect message boundaries -- not in the previous version (January).] d. Injong to draft the CFP (by end of May). [Ongoing] e. Stas to research using a modified IETF-like NoteWell re: IP issues and provide Injong with a researched IPR (by 5/13). [Done -- Internet2 has a standard one for WGs] f. Stas to send out `last call' for changes on the design document. [Done] Discussion of the UDT review Action Item: Stas hasn't had a chance to look at it yet and Gu is on vacation; concern whether we can implement the API on top of UDT. Steven hasn't had a chance to look at the most recent version; he has concerns about partial reliability and the ability to respect message boundaries -- not in the previous version (January). Stas is concerned that it is thread-aware and this would constrain applications to necessarily link with thread libraries and kernel support. Also, the API being object-oriented might be a hurdle. Steven agreed. Stas asked Chet if the e-VLBI code would have difficulties with this; Chet reported that David Lapsley had noted problems with UDT using the existing code (he would send David's comments to the list) but that the code would be changing in the future. Discussion of IPR; Internet2 has a standard ``Note Well'' statement for working groups that we can use. Stas sent a copy of this to Injong to review. Stas reported that he was giving a talk on Bulk Transport to Google in two weeks; they are preparing to use Internet2 to move data from libraries and are preparing for the problems they might be encountering. Larry noted that there are several groups that are beginning in areas touching on transport issues -- an Internet Draft is being developed by a colleague of Larry's (Fred Baker). He recommended the Design Document to Fred for review. Discussion of AQM schemes between Stas and Larry; Fred is concerned with keeping queues short. Larry is interested in getting the terminology nailed down before the potential NSF funding comes through so that everyone is using the same terms. Stas suggested that Larry request Fred to get comments in on the Design Document before it goes final; this way, if he has different terms he would like the community to use, he could start this discussion before the document goes final. Larry felt he would send comments by the end of next week. ACTION ITEMS: a. All (except Stas & Jeff Boote): Review the draft API Steven sent out (v.6.) and get comments to Steven. [ongoing; Chet will review beginning of next week and get comments to Steven by 5/20.] b. Injong to track the COMNET call for papers and keep the group posted when a date has been published. [ongoing] c. Stas to further review UDT Protocol as a starting point -- Steven will also consider the protocol with reference to his application/API needs. [ongoing] d. Injong to draft the CFP (by end of May) [ongoing]. e. Chet to send David Lapsley's comments on UDT to the list. Next call is at 12:00 p.m. on 27 May 2005 using the usual dial in numbers and pin (0111111). Call ended at 12:40 p.m. EDT.