IEPM

Status Report for SLAC/DOE IEPM Project

For March 2003

SLAC Home Page
Work completed or in progress since the last report.

Work Related to Analysis and Development

PingER

Warren obtained access to the PingER machines at Imperial (ic.ac.uk) and Wisconsin (wisc.edu) and installed a script to upload the PingER data to SLAC.

The PingER sites and Pacific NorthWest gigapop (PNW) and Brookhaven National Lab (BNL) crash because processes are remaining active and after some time no more processes can be spawned. Warren worked with Fabrizio to improve the process handling in PingER so processes would terminate if they exceeded 60 seconds.

The PingER machine at LANL was restored to service.

Warren worked with Les to update some beacons sites.

Warren worked with Frank Nagy and Maxim Grigoriev (FNAL) to create a CVS repository for PingER development.

MAGGIE

Warren, Les, Brian Tierney (LBNL), Vern Paxson (ICIR) and Andy Adams (PSC) submitted a pre-proposal on network monitoring to DoE for review.

Les discussed the MAGGIE proposal with THomas Ndousse and shared feedback with collaborators. We also discussed possible collaboration on Maggie with Rich Carlson of ANL.

IEPM-BW

Warren upgraded the IEPM-BW monitoring site at Georgia Tech.

Warren worked with a member of the network group and the University of Alabama at Birmingham (UAB) to analyze the results from the monitoring from Georgia Tech.

We worked with Intel and LANL to get delivery of 10GE NIC cards for the Sunnyvale testbed. As soon as we got them there ensued a week of frantic actyivity to make measurements. At the end of February the connection at StarLight had to be disconnected since the Juniper port was needed by the Tergrid. We discovered a problem with large windows (> 2MBytes) with bbcp and reported to the author.

Les started work with Fabrizio Coccetti on comparing the performance of various TCP stacks (HS, FAST, Scalable and New Reno) on production networks

NetFlow

Following a request from DoE for more information on characterizing ESnet and SLAC traffic, Connie and to a lesser extent Les worked with other DoE Labs to come up with a definition of how to characterize the traffic and started extneding the current SLAC Netwlfow traffic to make it suitable for public consumption.

Ipv6

Warren and Paola worked on resurrecting some of the IPv6 network at SLAC.

Web Services

Warren worked on a Web service front end for IEPM-BW monitoring sites.

Warren looked into the python implementation of SOAP in order to create a web service client for the Advisor tool.

Collaborations

We worked with NERSC to look into evaluating their passive monitoring tools, and compare with NetFlow.

Les and Warren met with Harvey Newman and Iosif Legran of Caltech to look into deploying Mona Lisa at SLAC.

Work related to Publicity and Information

CHEP

Warren and Les worked with collaborators at ICTP to write a talk and a poster describing the work on monitoring the digital divide. Warren also presented a talk on troubleshooting network performance for production grids.

Les put together a presentation for CHEP03 on a High Performance WAN Testbed: EXperience & Results. Also as an organizer for CHEP03 Les made last minute arrangements for the networking session.

HENP

Les & Warren put together and gave a presentation on Wide Area Networking at SLAC for the BaBar data distribution working group.

Awards

On February 27-28, 2003, over a Terabyte of data was transferred in less than an hour between the Level(3) Gateway in Sunnyvale, near SLAC, and CERN. The data passed through the TeraGrid Router at StarLight from memory to memory as a single TCP/IP stream at an average rate of 2.38 Gbits/s (using large windows and 9KByte "jumbo frames"). This beat the former record by a factor of approximately 2.5 and used the US-CERN link at 99% efficiency. With Caltech, CERN, and LANL we successfully submitted an Internet 2 Land speed record request. We also successfully submitted a CENIC On the Road to Ten Gigabit Award for: Biggest Fastest in the West proposal. As a result of the Internet2 Land Speed REcord awards we spent a lot of time wrting publicity releases and talking to reporters. Among other places the record was reported on the BBC, CNN, the Times of India, Nature and in Spanish, Portuguese, Japanese, Dutch, French and English. All good publicity for the DoE. As a result of questions asked we put together an FAQ on the record.

Reviews/Proposals/Meetings/Papers etc.

Following a request from Thomas Ndousse we set up a meeting with Werner von Braun, Brian Tierney, K Claffy, Ronn Rittke, and others at CIADA/NLANR to look into preparing a presentation for the LSN (inter department Large Scale Networking group) on the needs for network monitoring. WE will make a presentation in June.

Les attended the SciDAC meeting in NAPA and had useful discussions with the Rice and LANL people about the INCITE proposal.

Made final proof reading for paper to be published by Elsevier on the iGrid2002 demonstrations.

As part of a Caltech led NSF proposal we met with Tom West of CENIC to discuss National Light Rail plans.

Brian Tierney of LBNL and Les agreed to co-chair the next Protocols for Long Distance networks to be held in Chicago.

Digital Divide

Warren and Les met with Enrique Canessa of ICTP to discuss a "Round Table Discussion" meeting next November on the Digital Divide. We also worked with Tom Mead of SLAC PR group to put together a potential article on the Digital Divide for the CEDRN Courtier.

Created March 16, 2003. Last updated April 16, 2003.
Comments to iepm-l@slac.stanford.edu