1. Agenda bashing

2. Getting focus groups in line

The call began with discussion about focus groups. Chas reported that the first focus group went well. There was some question about getting notes written up. Russ agreed to contact Eric about next steps for this process. Discussion then turned to the next focus group and who should be involved on the interview team. Russ announced that his responsibilities had changed within Internet2 so he would be moving on. Matt agreed to continuing to participate. Chas asked for those on the call to think about who would be good for the next round of focus groups. He'd like to see folks who were more deeply involved with the tools.

Discussion then turned to what different tools were being developed as Internet2 projects like Internet2 Detective. Matt pointed out that Eric was key person involved with end-to-end performance tools. It was mentioned that there are other tools like the detective and the Abilene Observatory that are at different places in the Internet2 web space. Matt mentioned the E2Epi Tool list http://e2epi.internet2.edu/tools_list.html as a good resource list for performance and measurement tools. Chas reminded us of the MWe2eD tool survey http://middleware.internet2.edu/e2ed/public/survey/surveyhome.html.

3. EDDY Update

Chas gave an update on EDDY development. He said that the transport is coming together nicely. He pointed out that they were having some issues with capturing large numbers of events in XML with an apparent cap of 3,500 to 5,000 events per second. He said though, that on campus they feel that after tuning they should go from the 4 to 5,000 per second up to 8 - 10,000 events per second.

The next step will be to build an EDDY normalizer with optimized transformations to function at high speeds.

Chas said they have taken on another developer who is very good.

Jim will be doing two things next. He's working on an anonymization agent with the Dragnet team to change the key every week. This will be done in JAVA. It will be able to anonymize a couple fields in XML at 6,000 events per second and later optimized to 12,000 per second. He also will be testing the extendibility of the Common Event Record (CER). The CMU Architecture School is doing collections from their sensors in environmental domain. It has been very interesting. The plan is to to port it to Microsoft environment. This will include a high collection of small devices in non-IP network. Then the test will be moving from non-IP to IP to normalizers and transformational

Chas said they have the Dell machines all setup. So now we have horse power to move to higher rates. Now have grant from Sun should give another 9 machines.

He reported that the visualization developer will be able to have first visualization tool available soon -- topE for Top Event - this will show top talkers across backbone. It will be a Top Event view of a system to show real time views of what's happening. Later in the summer there will be archive agents.

Chas reported that everyone has been busy the past month with the Internet2 Spring Member Meeting sessions. He stated a goal to have a base ready to go for version 1.0 by September for folks to try out. Then locate some early adopters to work with this.

The session ended.