When

Starting Time: April 15, 2010 at 1:00 PM, US/Eastern

Dial In Information

If you only want to participate in the audio conference, you may dial in with the following numbers:
Call: +1-734-615-7474 (Please use if you do not pay for Long Distance) or +1-866-411-0013 (toll free US/Canada Only)
Enter access code: 0187242

Agenda

  • OGF Update (Martin/Jason)
  • GEC7 Update (Ezra)
  • UNIS Poster at the FMM (Marcos)
  • LS Scalability testing
    • Marcos
    • Ahmed
  • F2F Agenda

Attendees

  • Attending:
    • Marcos
    • Ahmed
    • Jason
    • Ezra
    • Martin
  • Apologies:
    • Andy Lake (ESnet)
    • Guilherme Fernandes (Udel)

Notes

OGF Update

Jason gave the OGF 28 update. Was held in mid March in Munich.

NM

Document is near last call. Needs last edits from martin (outstanding svn commits) and Jason. Will be handed off to other editors from NM/NMC/NML groups for final pass. Will go into WG last call. Then on to OGF process. Expecting to wrap this up before OGF 29.

NML

Group has finished one document. Bit of a lull now. Need to work on what is being termed the 'schema' document, but this is really an object model (e.g. we are not using schema langauge for this, outlining the objects. etc.).

After this there will be 2 other documents, one for schema and one for syntax. Both will take the above core object model, and bind to RDF (NDL) and XML (UNIS). Possibly something else for cNIS.

NMC

Result code progress. Progress on examples and some of the document too. Need more participation. Will have something ready for next OGF.

NSI

Lots of activity, not always in a positive direction. Some documents are close to being complete. Starting on protocol now (in addition to architecture, and use cases).

GEC7 Update

Ezra and Guilherme attended for UDel/LAMP. Guilherme gave a presentation in the I&M working group about schema. Some good feedback. Also relates to a call on 4/14 with the chairs of the I&M WG to see about adopting perfSONAR.

Saw a couple of demos, and talked to other measurement people (Purdue/HP, also Kentucky). Next steps will be having them integrate into the pS model (data or metadata).

Working on UNIS/RSpec (GENI language) integration. There is a practical deployment at UD (SNMP monitoring switches in GENI PoPs). Just passive measurement now. Next steps will be getting this into the experiment level. Setting up GENI software on a machine at UD to closer to this.

Martin bring up the conversations with the I&M group. We are not trying to get them to take UNIS as it is today, more as it will evolve. GPO would be able to push others to work to a larger and more encompassing solution. We will need to watch for pushback from the other measurement groups.

ESCPS (End Site Control Plane Service)

Also some recent work here - extend OSACARS work to the end host (lambdastation/terapaths).

UNIS Poster at the SMM

Marcos working on it. Still a draft, nothing to share yet (will share by Monday). Showcasing the use cases from the wiki. Asks about including GENI use cases. Martin says sure, as long as we note they are looking into it. Also some talk about how URNs work.

LS Scalability Testing

Ahmed: Will be writing some software to capture queries/timings form a live instance (via logging). Will then write software to 'replay' this to a test hLS/gLS. Will be working with Jason to reduce the number of transactions (read/write) in the code. Will get back into production infrastructure slowly (after testing). Also wants to look into caching, and how that can help.

Marcos: Started using PlanetLab. Wanted to test scalability of services registering to an hLS and hLSs summarizing. Deployed and working. Also getting VMstat info for Eric P at ESnet. After some analysis it was decided that nothing is CPU/IO/Memory bound. Some blocking noticed (in threads). Probably related to the dbxml operations.

F2F Agenda

Should talk about the same general things. No Video/Call access, jason will donate laptop for H323.

  • No labels