Status Report for SLAC/DOE IEPM Project
For August & September 1998
Les organized accommodation and travel for trip to PNNL for the ESSC meeting in mid September, prepared a talk on End-to-end Internet Monitoring, attended the meeting and presented the talk. In response to a formal request from the ESSC working group on international connectivity, Les reviewed a list of about 50 sites that they want to monitor to represent the ESnet programs. The results can be found in: http://www.slac.stanford.edu/comp/net/mon/essc-int.html.
Started making arrangements for travel & accommodation for the ESCC meeting in Chicago in mid-October.
Les reviewed XIWT SPSS proposal and made some suggestions.
Both Les & Warren attended the CHEP98 conference in Chicago at the end of August. Warren Matthews presented a talk and a poster at the CHEP98 conference in Chicago. About 20 copies of the poster were also sent to interested people. We started work on a paper to be published as part of the CHEP98 proceedings. Les organized and chaired a Birds of a Feather session on IEPM at the CHE98 meeting. Dave Martin is leaving HEPNRC to got to IBM, so we had discussions with Dave & Bill Lidinsky about HEPNRC's continued commitment to the IEPM project. Talked to Larry Price and arranged to give a presentation on PingER at the ESSC at the end of September. Talked to a lot of people at CHEP98 about Internet monitoring.
We have been providing assistance to a student at FNAL that Bill Lidinsky has working on generating baselines and alerts.
Discussed with Val Federov of ORNL the joint effort with SLAC on statistical modeling now that he is leaving ORNL. Val gave us another contact. We were hoping this would be driven from the ORNL end, so are awaiting to hear from the contact.
Made arrangements to attend a Surveyor meeting in Chicago in mid-October.
We met with NetPredict to go over their analysis of SLAC sniffer data. Charley Granieri worked with NetPredict to put together a Technology Transfer proposal to SLAC management.
BCR published "Why We Can't Compare ISP Performance - Yet". It has 4 charts which are from the PingER project and devotes about half a page to describing the results including mentioning Les Cottrell.
Warren investigated the idea of presenting an off-line demo of the WWW-CGI front end of the analysis results. Off-Line in the sense that it wouldn't require a network connection, but instead the pages would be served from a local web server on the laptop, all that is required is the latest processed data files and cgi scripts would have to be uploaded to the laptop. This project is not technically difficult, but more customization and localization is required than I had first thought, hence it will be a little time consuming to do it.
In order to improve the user-interface, it seemed reasonable to try to understand what are users are looking at. To this end Warren has begun to analyze the web log to see what are the hits on the program pingtable.pl
Surveyor is working well and taking data. Some of the surveyor nodes have been added to the nodes pinged from SLAC, so we can compare the performance reported by Surveyor to performance reported by Pinger.
Found some information from NLANR on the typical number of AS' between sites (see http://moat.nlanr.net/ASPL/) This will be useful as we design the NetMap procedures for showing traceroutes topologically.
We worked with Hank Nussbaum of Israel and other monitoring sites to refine the monitoring of Israeli hosts in light of increased blocking of pinging by Israel.
We have updated the Tutorial on Internet Monitoring to add more information on how ping packet loss may indicate Web performance and to show how the response time is related to distance and hop count.Back to Top
Revised 6 October 1998
Comments to firstname.lastname@example.org