Status Report for SLAC/DOE IEPM Project
For May & June 2000
Work completed or in progress since the last report.
New TimepingThere were some minor installation problems with the new version of timeping at HEPNRC, now fixed.
Warren worked on presenting the performance data as animated gifs. This makes it easier to see sudden changes.
Les released an updated beacons list.
SoundsWarren and Les have been working with some of the computer music people from Stanford to supply them with packet loss and round trip time data going back several years so they can use the data as inputs for pitch and other musical variables so we can hear how the performance has changed over time.
QoSLes worked on understanding US-UK QoS performance and created a web page to document his findings.
Rate LimitingLes worked with Mit to get his RA worked written up.
JASThe new servlet area at SLAC is under construction. The new graphing program will go into production soon.
Web LogsWarren and Les worked with the webmaster to understand problems with the logging on the web server that caused the weblog analysis code to fail.
TracerouteWarren installed an IRRD server at SLAC to be used to add the AS# to the nikhef traceroute.
IPv6Warren received email from a new volunteer wishing to be involved in the monitoring of the IPv6 6REN network, but the specified host is not routable. Warren is trying to work with the volunteer to find out where the problem is.
Missing DataThere has been a problem with getting new data into the graphing program at HEPNRC. This seems to be mostly fixed, but some links are still missing.
There was a minor glitch with data from Riken, but in all it has been a very good month for data collection.
We received email from the beacon site in Belgium and the Motorola site requesting they no longer be included in the monitoring. In the case of the Belgium site we have already replaced them with a new beacon site. The Motorola site may return, the contact is currently reviewing the status with the administrators.
New SitesCaltech became the latest site to join the pingER project.
Two sites in South Korea sent email requesting information on running PingER.
Throughput to CERN/IN2P3Gilles Farrache had observed less than expected throughput transfering data from SLAC to CERN. Warren examined the TCP flows using the BaBar FTP program using TCPDUMP, TCPtrace and xplot. Per flow loss rates were found to be <1%. However, it was also found that the machine using to send the data wasn't powerful enough and the poor rate was due to insufficient memory on the computer and not a network issue.
IPEXThe XIWT Internet performance measurement work will receive some further funding, and some of the monryu will come to SLAC to fund a student to assist in the development and analysis.
SupportLes worked with a site in Australia to install ping_data_plot.pl
Enquiry from Alcatel regarding interpreting SLAC-analysed surveyor plots.
Enquiry from Pulse for help with TCPTRACE.
Enquiry from an individual enquiring about network performance.
TalksWarren has been invited to make a presentation to the HEPCCC group about WAN performance.
New RAImran Q Sayed will start mid-August
Mit, Faisal and Yvonne have finished.
TrainingWarren attended the Interconnecting Cisco Network Devices (ICND) to learn more about Switches and Routers.
Back to Top
Revised 27 May, 2000.
Comments to firstname.lastname@example.org