Status Report for SLAC/DOE IEPM Project
For August 2001
Work completed or in progress since the last report.
IPEX/NIMIThe contract between SLAC and CNRI is almost complete and SLAC will soon be able to hire a student to work on IPEX analysis and data distribution. Warren worked with Andrew Adams, chief developer of the NIMI software, to understand aspects of the NIMI architecture required by IPEX.
INCITEWarren contacted the INCITE collaborators at Rice University and LANL about setting up PCs to install PingER and NIMI and about obtaining the Chirp software for deployment at SLAC.
Web100Doung Chang updated his Web100 page.
GridWarren is working with Adil Hasan, the BaBar database administrator and the local expert on the Globus toolkit to develop and LDAP schema and deploy an LDAP server to make monitoring data available.
Warren also worked with Warren Smith (Nasa/GGF) to understand his LDAP schema and access the test database to include SLAC details.
High throughput/QBSS Les and Jerrod worked on automating high network and file transfer throughput performance measurements. Les made file transfer measurements with bbcp to several sites (Caltech, CERN, Dresden, DL, LBL, IN2P3) and created a web page on Bulk file transfer measurements which quickly grew to a size that Jerrod is now reorganizing the page. Worked with the bbcp developer to add some useful measurement features (shorter measurement reporting intervals, source from /dev/zero, destination to /dev/null) and to help debug it. Jerrod wrote a script to automate testing of bbcp compression, but ran into a bugg that prevented further progress.
Paola set up and configured 2 Cisco 7200s to make up a new QBSS test bed at SLAC. Validated that the QBSS code point is transmitted fro CERN and DL to SLAC. Made measurements on the effect of QBSS and documented.
Succeeded in getting access to a test machine at IN2P3, and with it broke the 100Mbps barrier for TransAtlantic traffic.
Spent a lot of time validating high performance measurements to see how they behave with different OS', with different measurement durations, etc.
Passive MonitoringWith Connie Logg Les looked at analyzed Netflow large (> 10MBytes) flows. Les & Connie also gathered and analyzed packet size and throughput distributions from netflow data. Compared netflow with iperf throughput. There is good agreement apart form at high rates. The above information is included in Characterization of the Traffic between SLAC and the Internet and will be presented at CHEP01.
VoIPWarren has been working with Texas A&M University and the Internet2 VoIP Working group to take part in a demonstration at the Internet2 Members meeting.
Netflow information on usage of the VoIP systems is being gathered.
- Imran has been developing a port of traceping.
- Warren installed the pathrate tool.
- Warren installed the Packet-Loss Warning program from TRIUMF. However, there is an unresolved problem with the graphing compared to the original deployment at TRIUMF.
- Warren looked at installing the Enable Tool. Installation was thwarted by a python error and the lack of expertise to resolve it.
TracerouteWarren reinstalled and the whois server at SLAC, and Les modified traceroute to use it once again.
Missing dataThe Univeristy of Melbourne, INFN, APAN Korea and ARM sites have all been down all month. Repeated emails have failed to resolve the issues.
We found about 5 sites that were blocking pings, and then contacted various people to get alternatives.
Problem assistanceWe assisted the Technical University of Dresden im measuring the performance to SLAC. To do this Les got an account at TU-Dresden and installed iperf, tcpload etc. See http://www-iepm.slac.stanford.edu/monitoring/bulk/dresden.html
IEPM WebSiteWarrren added a What's New section to the IEPM website.
Les added outreach information to the IEPM web site.
DoE/ESCC/ESSCLes started preparing a presentation on ESnet monitoring, for the ESnet Review which is scheduled in Sante Fe on September 10th.
Les created a yearly report on IEPM accomplishments for DoE.
Les provided SciDAC award information to SLAC's publicity office.
HENP/CHEPWarren, Connie and Les submitted several papers for the Computing in High Energy Physics Conference in Beijing at the start of September.
IETFWarren attended the IETF conference. His trip report is available.
CAIDA/NLANR/PAM/NIMIMargaret Murray, CAIDA Technical Manager, emailed Warren and Les about including the Performance Maps in the Internet Atlas Gallery
Les is on the organizing committee for the PAM 2002 meeting which will be in Colorado.
Les set up a meeting between Vern Paxson and SLAC security folks to look into deploying a passive monitor at SLAC.
Internet 2Les is discussing with Guy Almes of I2 a possible collaboration on a proposal, and also providing mutual access to high network performance machines for throughput testing.
GridsLes created a monitoring activity report for the PPDG.
SC2001 Les put together a new draft of the panel for SC2001. Also working with Tony Johnson to provide a real time display tool (UTH) that will allow us to show performance in real time.
IndustryWarren met with Nai-Ting Hsu and Mark Crane from Network Physics to discuss Network Monitoring and Technology.
Working with QWest to try and get an OC48 to SLAC for the duration of SC2001. Also working with Sun on a possible high throughput testbed. Les wrote up an intial test plan.
Created 3 July 2001.
Comments to firstname.lastname@example.org