Ben Teitelbaum opened the meeting by announcing two themes: constructive self-criticism (what the working group is doing vs. what it should be doing), and bridging the gap between the QBone vision and the QBone reality. The four agenda items were a QBone update, the upcoming QBone workshop, the upcoming QBone interoperability event (QCon), and the following day's joint QoSWG/QBBAC meeting.
Ben announced that the QBone Architecture is now ready to go to v1.0. Once the document is at v1.0 the next step will be to assemble a team of QBone "czars", or human bandwidth brokers, for the Phase 0 deployment. He also called the group's attention to "a certain amount of stagnation" in practical preparations for the QBone launch. Russ Hobby noted the that at this stage it is difficult to know what problems to focus on, and was seconded. Ben agreed, noting that there are different problems for different users with different equipment. This led to a short discussion of miscellaneous snags slowing down the QBone deployment. Ben then resumed his update by annoucing that there would be "three major sources of agitation" in the next few months: the QCon, the initial Abilene deployment, and "the big big QBone workshop". A large part of the big big workshop will be discussions of implementation experiences, which hopefully will motivate attendees to pick up the pace, so as to have some experiences to relate.
The discussion then moved on to planning for the QBone workshop. Ben contrasted the upcoming workshop with the QoS98 workshop. QoS98 was devoted to DiffServ background plus "figuring out what the problems are". The workshop achieved a consensus, which was published in the QoS98 report. This time, the object will be to get a read on where we are and where we are going with deployment, and to provide a forum for bragging and complaining. It remains an open question if it will be possible to announce the availability of a service at the next QBone workshop. The plan is to have some testing success stories to tell. Rüdiger Geib encouraged the group to keep the focus of the event narrow so people don't start thinking they have to reinvent the Internet in order to get involved. Ben restated the purpose of the workshop as to be a research forum, providing an understanding of the QBone project, and in particular an understanding of what one has to do to participate in it.
Rüdiger commented that especially as the working group had done a lot since the last workshop, it is important that this workshop not repeat things done at the last one. In response to Larry Dunn's suggestion, the group made the following list of things accomplished since QoS98, or expected to be accomplished by the next workshop.
Next John Streck reported on planning for the QCon. He has contacted several vendors (among them Cisco, Nortel, 3Com, and IBM), with more to come. The QCon document is ready to go to the vendors. He encouraged the group to keep expectations low for the early stages of interoperability testing — "The methodology is not that complex, but to pull it off is gonna take time." Many of the vendors don't want to do "an open bakeoff"; there are trust issues involved in interoperability events.
Finally, Ben gave an update on bandwidth brokers and IETF work on using IntServ for inter-BB communication. Sue reported that the work of the QBBAC had gotten bogged down in details over the last few months. While the intent had been to allow lots of variety in intradomain implementations and little variety in interdomain implementations, no one likes the answers, such as Diameter, that the current requirements document seems to point to. For this reason the group has had to retreat from specifics back to requirements, in order to find out how deep the disagreement lies. This led to a discussion of the state of the BBReq document. It was noted that "One of the big problems is what happens if you lose connectivity — what happens if inter- or intra-domain links fail?" Larry Dunn commented that with respect to this issue there is a tension between desire for simplicity (and resultant ease of consensus), and usefulness, and that as a result "this has been swept way under the rug". There was general agreement. Sue Hares noted that "You've come right to the heart of the matter: what happens when it breaks?" This led to a review of the planned phases of BB deployment, after which the meeting was adjourned.