Status Report for SLAC/DOE IEPM Project
For October & November 2000
Work completed or in progress since the last report.
QoSWarren and Les met with the team from Daresbury Lab to create the framework and develop an experimental program for QoS testing. A full report is available at DL/SLAC network measurement meetig, Nov 2000.
JNFlowThe JNFlow application monitoring has broken, probably due to changes in the SLAC LAN. Warren is working with the analysis code developer to bring the system back up when changes stabilize.
Web100Warren installed the pre-alpha software from the web100 project, and an application is under development. Les discussed and got a demo of how to use Web100 from the NASA contact at SC2000.
RoutingWarren looked at routing to CALREN sites via Campus. It may provide improved performance to LBNL and LLNL during the ESnet transition. Les worked with BaBar, LLNL, LBNL and ESnet folks at the ESCC meeting to help define ways to temporarily improve things.
IPv6Warren has been working to bring up IPv6-aware NFS, NIS and other application on the new Solaris 8 machines. Warren has been working on a Field Work Proposal with input from ESNET, FNAL, ORNL, and CRC. Les discussed submitting such a proposal to MICs with Rchard Carlson, George Seweryniak and Thomas Ndousse.
Traceroute etc.Responded to 3 sites who mistakenly thought our reverse traceroute server, or traceping was a port scan.
Missing dataThere were glitches with data collection from Stanford, CERN and Monash this month. Problems remain with collecting data from Caltech, DoE/MICS, and the University of Wisconsin, but I am now convinced this is due to a problem at SLAC or the local host machine.
SC2000At the SC2000 SLAC/FNAL booth (see SLAC and FNAL at SC2000), PingER/IEPM had a panel with 3 monitors with real-time interactive displays of the monitoring. Les helped set up the booth and networking and attended the panel to provide assistance to visitors. There was considerable interest. Les also put together a web page on SC2000 to SLAC high thruput performance measurements made at SC2000, see SC2000 bulk thruput measurements to SLAC.
ESCCWarren presented an update on PingER results at the ESCC Fall meeting. Les made a presentation on High thruput file replication (see High thruput file replication) and took some rough notes of areas of special interest, see: http://www.slac.stanford.edu/grp/scs/trip/escc-nov00.html
PAM 2001Warren submitted an abstract for the Passive and Active Monitoring conference. See Passive and Active Monitoring on a high performance network.
ProposalsLes worked with Vern Paxson to put together a draft proposal on a proposed ESnet measurement infrastructure. Les is working with Rice and GATEch on possible joint proposals.
Les met with the chairman (Nobel laureate Burt Richter) of the International Union on Pure and Applied Physics (IUPAP) to review how to provide information useful to promote improved Internet connectivity for developing nations.
UKLes has been working with UK folks to understand increased RTTs. These were caused by bbftp file transfers. Helped coordinate efforts to request a bigger pipe between JAnet and ESnet (at the moment it is restricted to 10Mbps).
EnquiriesWe have received a number of emails for futher information on projects on the group web site. Topics include trans-atlantic jitter, usign TCPDUMP, information on Internet2, obtaining PingER software, packet duplicates and performance on ESnet. It's great to know our work is proving to be so useful to so many people.
Back to Top
Revised 22 November, 2000.
Comments to email@example.com