Dec. 19, 2006: Difference between revisions

From ReddNet
Jump to navigation Jump to search
No edit summary
Line 51: Line 51:
*Micah will recover and put on line the old IBP spec or, if that can't be located, do what he can to replace it.
*Micah will recover and put on line the old IBP spec or, if that can't be located, do what he can to replace it.
==Some Notes==
==Some Notes==
*Discussion of software specifications for LN/REDDnet sofware
**We talked about the need to have specs for various interfaces so that different implementations can retain interoperability and people can be confident that what build on won't change out from under them.
**It's problematic because we have a schedule that we need to meet and achieving agreement on specifications can be time consuming.
**The most immediate challenge is libxio, which is lacking in various important respects, such as the ability to deal with the parity mappings that LStore does and the lack of caching.
**Parallel disucssion of need for (and current lack of) documentation and adequate comments of much of the LN code.
**Conclusions:
***Alan and the ACCRE group want to get libxio replaced as quickly as possible. So the compromise process is a) put the rough spec they will create to guide their reimplementation on-line, b) implement a new version of libxio, c) document the code d) go back to the rough spec and modify it in the light of the new codes/comments. The idea is that the spec that results at stage d) will be our initial cut at the community spec. and we'll work toward and agreed upon a community spec.
***
*We agreed that we wouldn't try to go for a demo at TRIDENT 2007. We're overloaded already.
*

Revision as of 13:37, 19 December 2006

Coordinates

  • Time: 10:30EST/9:30CST
  • Conference #:510-665-5437 Meeting ID:7333

Attending

  • Hunter Hagewood (Nevoa)
  • Micah Beck, Terry Moore (UTK)
  • Dan Engh, Paul Sheldon, Alan Tackett (VU)
  • Mike Coffee, Diana Gunter, Shane Jenson

Current Agenda

Application Communities

Deployment and Operations

  • Review of the current deployment schedule
    • Is there any update on this schedule? A timeline of what will be deployed where and when?
    • There are demos planned for the beginning of March (AV meeting in Washington at least, maybe OSG meeting in SD as well).
  • Discussion of alternative strategy for investment in REDDnet deployment - The general idea is to buy more, but perhaps less well provisioned nodes, and try to deploy them to more places where application groups will fully populate the rack with storage. (Terry)

Technology: Hardware, Software

Organization and Funding Opportunities

  • NSF SDCI (Due. Jan. 22)
  • NSF STCI(Feb. 8th)-- What kind of opportunity is this (Micah)?
  • NSF PIRE (Feb. 28th)-- Paul: Any word?

Events, Education, Outreach

Action Items

  • Terry will get the LN wiki up and going (http://www.logisticalnetworking.org)
  • Alan will share the rough spec he's going to develop for libxio before he implements
  • Micah will recover and put on line the old IBP spec or, if that can't be located, do what he can to replace it.

Some Notes

  • Discussion of software specifications for LN/REDDnet sofware
    • We talked about the need to have specs for various interfaces so that different implementations can retain interoperability and people can be confident that what build on won't change out from under them.
    • It's problematic because we have a schedule that we need to meet and achieving agreement on specifications can be time consuming.
    • The most immediate challenge is libxio, which is lacking in various important respects, such as the ability to deal with the parity mappings that LStore does and the lack of caching.
    • Parallel disucssion of need for (and current lack of) documentation and adequate comments of much of the LN code.
    • Conclusions:
      • Alan and the ACCRE group want to get libxio replaced as quickly as possible. So the compromise process is a) put the rough spec they will create to guide their reimplementation on-line, b) implement a new version of libxio, c) document the code d) go back to the rough spec and modify it in the light of the new codes/comments. The idea is that the spec that results at stage d) will be our initial cut at the community spec. and we'll work toward and agreed upon a community spec.
  • We agreed that we wouldn't try to go for a demo at TRIDENT 2007. We're overloaded already.