NLANR/DAST Monthly Update, October 2002

(Unless otherwise noted, all persons mentioned in this report get part or all of their funding from the NLANR/DAST cooperative agreement.)

a. Activities related to network tool development for end users

• Research assistant Kevin Gibbs has taken on the main responsibility of Iperf and has hit the ground running. Addressing a list of requested features procured the previous month, Gibbs has tackled the highest priority items of two-way testing and summing of multiple thread Iperf runs. In addition, bugs were corrected from the recent updated port of Iperf. A demo Alpha version of the latest completed Iperf is planned to be used and demoed on the floor at SC2002.
• Steve Engelhardt, Tanya Brethour, Richa Hingle and Feng Qin continue to develop various aspects of the Network Performance Advisor, formerly known informally as the "Framework". Engelhardt and Qin have been focusing on data collection, authorization and security, while Brethour and Hingle are focused on the user interface and analysis engine. Jim Ferguson is providing overall coordination of the Advisor effort, as well as being chief liaison with Internet2's testing project. The team has focused on a version of the Advisor with an "expert" interface to be shown on the show floor at SC2002 in Baltimore. The GUI has prompted much extra work to make it look like something that users and engineers would like to use. This goal has initiated some integration work that had been neglected up to this point of the project.
• Research assistant Yanli Tong has made excellent progress (with the shepherding of John Estabrook and Ferguson) on her project to adapt the NLANR/DAST auto-tuning FTP code for further impact on the community. The plug-in for GridFTP works, and can be distributed. The team is assessing some changes in the specification for GridFTP that affects parts of the code and plug-in. It is hoped that these spec changes will not have a major impact on the work done so far, but that is not clear yet.

• An Access Grid meeting regarding the NLANR Multicast Beacon was held with collaborators from Lawrence Berkeley Lab, specifically Deb Agarwal and members of her group there. Tony Rimovsky and Mitch Kutzko are the main Beacon people at NLANR/DAST. The planned port of the Beacon client from Java to Perl is now cancelled, and the redesign (2.0) and new development will take place directly based on the current code base. The cancellation of the port was decided because (a) it is not critical to current operation of the Beacon, and (b) the redesign and advancement to version 2 can begin forthwith. The Beacon will use RTP in version 2, and Kutzko has been busy familiarizing himself with this protocol.
• See GridFTP work above also
• John Towns has participated in regular development activities related to various aspect of the TeraGrid/DTF project as co-lead of the User Service Working Group and as a member of the Grids, Networking, Applications and Accounting Working Groups. In addition, Towns made significant contributions during the TeraGrid Implementation Planning Meeting held Oct 28-29 in Chicago, IL during which much of the execution environment for TeraGrid resources was defined.
• John Towns as provided input and guidance on support-related issues for the GRIDS Center/NMI project. New models for documentation and support for the growing software component set are under development in order to appropriately scale.

• Greg Cole and some staff borrowed temporarily from other projects have made great headway on the NextINet project for information push to the network applications community. This site is based on community-development efforts done by Cole and Bulashova before at Friends & Partners. See http://www.nexinet.org for the current look. This has not been advertised yet, as a critical mass of content has not yet been achieved. Much of the functionality of signing up for information push, customizing, search, etc. has already been implemented.
• Steve Engelhardt worked with the Berkeley-Illinois-Maryland Array (BIMA) project more this month, but mostly in a 'cleanup' mode as he transfers the work he has done for this community over the past year for them to maintain and expand.
• John Towns, and Jim Ferguson participated in the Global Grid Forum 6 meeting Oct 15-17 in Chicago, IL. Activities of the Grid User Services Working Group progressed. Towns also participated in the Applications and Testbeds Working Group and in the further development of the Production Grid Management Working Group.

• The BIMA suite of codes (see above) astronomy community applications.
• No other similar codes in October.

• Jim Ferguson traveled to the Korean Institute of Science and Technology Information (KISTI) in Daejeon, South Korea on the invitation of Dr. Ok-Hwan Byeon, head of the High-Performance Research Network division. Ferguson gave three talks in 3 days at KISTI, the first entitled "Supporting high-performance network Applications in a Grid Environment" given at KISTI's Grid High Performance Networking Symposium. The other two talks for members of KISTI's staff and other invitees were on NLANR/DAST and Web100. During his visit, Jim also met with Sangsan Lee, head of the Supercomputer Center of KISTI. The trip was paid for by KISTI.
• Jim Ferguson attended the Internet2 Fall Members Meeting in Los Angeles, Oct 27-30. Jim gave a talk on the Network Performance Advisor, spoke with NSF program manager Greg Monaco, and made contacts with several members of application teams, including some AmericaView participants in Texas and Alaska. Ferguson talked extensively with the End-to-End team from Internet2 on shared interests, especially in the meshing of the Network Performance Advisor with the I2 E2E piPES project.

• Ronn Ritke (SDSC, NLANR/MNA), John Towns and Jim Ferguson continue to develop ideas for cooperative ventures in the NLANR program. Both sites are eager to cooperate on projects. It appears that the Observatory Project is not ready for collaboration with NLANR now, as they did not get funding. NLANR/DAST and NLANR/MNA are still pursuing the Gemini project, as well as an information collection project, starting with a set number of common network problems, worked out in a systematic fashion. The NextINet project is also going to include NLANR/MNA involvement so that the community development efforts cover all of NLANR's collective efforts and communities.