--------------------------------------------------------------------------- From: "David Shealy" Date: Tue, 3 Aug 1999 16:17:26 -0500 Hans-Werner: I have read both drafts of the Measurement & Analysis and believe these reflect my understanding of important topics covered at this workshop. I personally would like to see the last-mile concept defined...I believe this term means different things to different groups of people...we have local loop issues (connectivity from telecom pop to campus demark; we have campus connectivity from demark over campus backbone to buildings; and we have building switches/hubs to wall outlets; and finally from wall outlet to CPU (cable, NIC, software drivers,OS). I've heard the last 2 connectivity issues called the "LAST-INCH" problem...Some campus telecom organizations are only concerned with backbone and to building switches and other campus telecom are concerned with campus backbone to wall outlet and other campus academic computing units are concerned with the actual performance users get on live networks. I believe it would help high performance networking community if these proceedings made it clear that all of these different components AFFECT end-to-end performance and one needs to deal with each one...the weakness link can kill benefits of others strengths.....I recall much of the discussion at workshop talking about these various things...how to setup TCP stacks....building AMP-type boxes to measure high performance network performance from desktop to desktop...so, I believe we acknowledged the overall scope of problems of setting up end-to-end HPN....I personally believe the phrase "last mile" is too simplified... Otherwise, I enjoyed very much the workshop and I am looking forward to setting up AMP type boxes at UAB campus....I understand that you have been communicating with both Landy and Jill about possible collaborations with UAB...a CS faculty member at UAB with good networking background is interested in working with NLANR on a campus deployment of AMP-type boxes. Let me know if there is anything I can do to further promote successful measurements and analysis collaborations with UAB... David David David L. Shealy, Ph.D. Chair, Department of Physics Director, Academic Programs Information Technology 310 Campbell Hall, 1300 University Blvd University of Alabama at Birmingham Birmingham, AL 35294-1170 Email: dls@uab.edu Voice: 205-934-8068; FAX: 205-934-8042 ---------------------------------------------------------------------------