Action Items

• Complete defining the features of the pilot
• Arrange for a BoF session at Internet2 Fall Member Meeting

Pilot Progress

Chas presented an update on what's going on in the pilot right now. Russ would be working on scenarios for passive and active measurements. Ken directed Chas to the IETF and other groups that are working on diagnostics, in particular security folks. One area of particular interest is how to provide non-repudiation for distributed denial of service (DDOS) attacks.

Next week Chas will be in Pittsburgh and will use that time as an opportunity to arrange an in person meeting with Renee Shuey and her staff at Pennsylvania State University if possible. There was brief discussion about passive and active measurement.

George and Ken had heard interest in this approach during meetings at NSF. Russ commented that the E2Epi Technical Advisory Group (TAG) had included passive & active measurement integration in their E2Epi road map for the next three years. He said the E2Epi TAG recognize the work of e2e Middleware Diagnostics.

Chas said that the two developers have been busy. They have made an initial analysis of netlogger and are now looking closely at AirCert. AirCert appears to have better possibilities of early use for the pilot than NetLogger which would require more coding and an API as well. They are looking at the two paths to see which they'll focus more effort on first. They plan to build the event record to use sub-schemas (written in XML) for each of the diagnostic areas (networking, security, middleware, and systems). AirCert is looking like the better choice and Chas knows the authors of the application. AirCert does have some rough edges. It took a while for the developers to compile and get it up and running. The documents are lacking. A decision is planned by close of business June 18 whether to build our own tool with NetLogger base or to continue with AirCert.

Russ had two questions: 1) Where does the data go at the end of a flow; and 2) Who is doing the work. Chas replied that the data will be pushed back to a central repository, a MySQL database. He said the work is being done by the two developers, Jim and Ryan. Chas noted that other universities have expressed interest in joining in the work. Chas said that the next milestone will be to complete defining the features of what we will produce. We need to be fleshing out the work plan for the next three or four months to have something by the end of September for the Internet2 Fall Member Meeting.

There was some discussion about similarities of work in the different areas: networking, security and middleware. Chas said that Ken had asked if the work should be consolidated or not. Should there be a framework to support different areas -- it may be the same but need different operational structure for each. Important things are to keep from re-inventing the wheel, keep it simple so that it won't take too much time to build or be too complex to use. Russ suggested getting representatives from the different areas together to explore the similarities. Chas agreed to the idea. It was recommended that there should be a session at the Internet2 Fall Member Meeting to bring together representatives from the areas. Last comment was that we are all curious of how the recommendations of the consultant for Internet2 will impact the project.