Meeting-Zurich-13Apr05

From GEANT2-JRA1 Wiki

[edit]

Attendees

Nicolas Simar (DANTE)
Andreas Hanemann (DFNm)
Roman Lapacz (PSNC)
Szymon Trocha (PSNC)
Loukik Kudarimoti (DANTE)
Jeff Boote (Internet2)
Jerome Durand (RENATER)
Martin Swany (UoD)
Olav Kvittem (Uninet)
David Schmitz (DFNm)
Roland Karch (DFNe)
Stephan Kraft (DFNe)
Maurizio Molina (DANTE)
Thanassis Liakopoulos (GRNet)
Janos Mohacsi (NIIF/HUNGARNET)

[edit]

Agenda / Actions

The general goal would be to prepare for development phase in terms of final design of messages, identifying ways of integration of SNMP data sources, choosing technology and splitting the work. The order of items is a subject to discussion.

  1. To discuss and take final decision regarding the technology to use
    • GLOBUS Toolkit
    • Axis
    • Action
      1. Use Axis for prototype
  2. Discuss licensing (Nicolas to provide some more info according to his last e-mail)
    • Action
      1. Wait for legal point of view. We can work and start coding within prototype provided a code is not published outside.
  3. Identify next steps
    • Action to split prototype into phases:
      • Phase I
      • Phase II
      • Phase III
  4. Discuss and finalise message content. The goal is to work and agree on the content of document for area 1
    • list of messages and terminology of messages consistent with GFD
    • Lookup Service messages
    • Measurement Point messages
    • Measurement Archive messages
    • Actions
      1. Use UDDI for prototype and look at eXist Open Source native XML database for future.
      2. Put technology evaluation on Wiki.
      3. Not to use metadata for phase I of prototype to specify which data is available in Measurement Archive and take decision how to proceed at phase II.
      4. Use WebServices for data plane for Measurement Archive.
      5. Use pull model for communication between CL and Measurement Archive.
      6. Use push model for communication between Measurement Archive and Measurement Point.
      7. For prototype stage capacity will be hardcoded together with topology data on the client side.
      8. For prototype the request to Measurement Archive will include:
        • Start and end time.
        • Type of aggregation: min, max, avg, last.
        • Resolution.
      9. For prototype the response from Measurement Archive will include:
        • Metadata of the request.
        • Data (start and end time, value).
  5. Visualisation of prototye (DFNm)
    • what to show
    • how to present
    • how the visualisation will access data
    • Actions
      1. Topology data will be entered manually for prototype.
      2. CNM acts as a cache of data. Data will be populated from Measurement Archive to local database with the use of simple command line application.
      3. CNM will use WebServices to access data from SONAR.
      4. D1.4 database schema is for visualisation tool only. Subsets could be reused later on for prototype phase III or later (e.g. for Measurement Archive, Topology Service).
      5. DFNm to contact with NRENs which agreed to provide RRD data to start working on topology visualisation and data access.
      6. DFNm to prove the schema with the use of SONAR metrics.
  6. Integration of IPPM data into next phase of prototype (Stephan)
    • Action
      1. Integration postponed until phase III. Could create Measurement Archive around IPPM database then.
  7. Integration of MP and cricket/RRD
    • Actions
      1. PSNC to explain MRTG configuration.
      2. LK to explain Cricket configuration.
      3. Start working on Cricket/MRTG migration.
      4. NRENs willing to provide data:
        • Uninet / NEMO
        • DANTE / Cricket
        • Hungarnet / Cricket
        • GRNet / MRTG
        • GARR / MRTG
        • RedIris / MRTG
        • Abilene / Cricket (?)
  8. Format of messages (Martin)
    • Actions
      1. MS to send presentation to the list.
      2. LK to update prototype document according to discussion.
      3. MS, LK, RL to work on schemas for Lookup Service. Start filling prototype document chapter with format of messages.
      4. MS, LK, RL to work on Measurement Archive response and request format.
      5. MS, LK, RL to work on Authentication Service message formats.
  9. AS messages (Maurizio)
    • Action
      1. MM to work out requirements for JRA5 (area 3 work).
  10. Discuss TNC2005 full article
    • Actions
      1. Incorporate parts from e2emon paper. Proposed structure:
        • Introduction taken from abstract and e2emon. (AH)
        • 3 layer description. (AH)
        • Services description. (RL, ST)
        • Multidomain use case. (LK)
        • Information about prototype and project schedule. (NS)
        • Related work. (JD)
        • Conclusion. (EB)
  11. Split the development work into pieces and allocate them to partners/people
    • Actions
      1. Decision postponed but should be taken within 2 weeks.
      2. People should familiarise with Axis and then decide how to split the work.
  12. Discuss and take final decision on LS data structure - NOT DISCUSSED

Back to JRA1 events

Retrieved from "http://wiki.perfsonar.net/jra1-wiki/index.php/Meeting-Zurich-13Apr05"
Personal tools