Changes between Version 6 and Version 7 of _GIMI_NextSteps


Ignore:
Timestamp:
05/04/12 17:03:03 (12 years ago)
Author:
hmussman@bbn.com
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • _GIMI_NextSteps

    v6 v7  
    88 - After status call with Mike and Max on 4/18/12:
    99 - After status call on 4/25/12:
     10 - After status call on 5/2/12:
    1011
    1112== 1)  Status calls and reviews ==
    1213
    13  - on 4/25:  when is next status call?  5/2
    14  - on 4/25:  when is design review?  not 5/2;  week of 5/7?  week of 5/14?
     14 - on 5/2:  when is next status call?  5/10 (Thurs)
     15 - on 5/2:  when is design review?  Thurs 5/17 in Cambridge?
    1516 - ***  marks items that will be important to review on the next call
    16    - 3)  *** Architecture of GIMI tools for GEC14 tutorial
     17 - +++ marks items for design review
     18   - 3)  +++ Architecture of GIMI tools for GEC14 tutorial
    1719   - 4)  *** Introduce XML messaging service:  required for OMF messages, used to orchestrate services
    18    - 7)  *** Steps towards GIMI testing:  start by forwarding all code and plans for GEC13 demo;  by 4/25?
    19    - 8)  *** Steps towards GIMI tutorial at GEC14
     20   - 7)  *** Introduce IREEL service
     21   - 8)  +++ Steps towards GIMI testing
     22   - 9)  +++ Steps towards GIMI tutorial at GEC14
    2023
    2124
    … …  
    3033
    3134
    32 == 3)  *** Architecture of GIMI tools for GEC14 tutorial  ==
     35== 3)  +++ Architecture of GIMI tools for GEC14 tutorial  ==
    3336
    3437a)  (Mike)  need to arrive at nominal configuration ASAP
    3538 - on 4/18:  good discussion (see below);  still need a clear and complete description soon;  when?
    3639 - on 4/25:  discussion as noted below;  still need a clear and complete description soon, and a full design review;  when?
     40 - on 5/2:  discussion about MAP as noted below;  design review likely on 5/17.
    3741 
    3842a')  on 4/20:  issue:  how to configure MPs/MCs?  from EC to RC in  MP/MC?
     43 - on 5/2:  if use IREEL to provide MAP functions, can 1st be used to configure the MPs/MCs, i.e., use IREEL and EC to configure the GIMI services, as if configuring a complete experiment.
     44 - on 5/2:  later, 2nd, consider extending IREEL to configure the entire experiment, i.e., interface with Flukes/OMNI, etc., to aquire resources.
    3945
    4046a!'')  on 4/20:  issue:  is there a MAP on common node?  at least to preview data?  is there a MAP in user workspace?  what?
    … …  
    4955 - on 4/25:  for 1st, expect MAP only bin user workspace
    5056 - on 4/25:  what is 1st MAP service?  what functions are provided?  can these be done using iREEL and r?
     57 - on 5/2:  1st plan for MAP service:  base it on iREEL 2.0, installed as a public service;  where installed?  not user workspace;  not BBN;  UMassAmherst?  RENCI?  NICTA too far away. 
    5158
    5259e)  movement of data from MC (OML server) to MAP service:   direct?   how?   via iRODS, i.e., MC to iRODS to MAP?  how?
    5360 - on 4/18,  per Shu, and discussion:  could login to MC node;  establish first transfer from iRODS client on MC to to iRODS with iput;  use  iinit, so iRODS client remembers password;  then periodic transfers with no human intervention;  another push protocol, from MC to iRODS.
    5461 - on 4/18,  per Shu, and discussion: next, can install iRODS client on user workspace, then pull data back to user workspace for analysis and presentation, and then push back to iRODS for archive
     62 - on 5/2:  must be able to pull data from IRODS to IREEL for MAP functions.  how?
    5563 
    5664f)  1st:  common node in slice;  2nd:  common node in user workspace (out of slice)  (is this beyond GEC14?)
    … …  
    7684 - on 4/18:  expect release of first OMF6.0 by mid-May, with firm RC and skeleton EC;  expect updated EC by mid-June;  is this too late for GEC14?
    7785 - on 4/25:  need to start with an earlier version of OMF, e.g., OMF 5.3
     86 - on 5/2:  need plan for using OMF 5.3 (or 5.4).
    7887
    7988== 5)  Tools in user workspace  ==
    … …  
    8695
    8796d)  (Ilia) FLUKES optional to create (or observe) rspec
     97 - on 5/2:  how does this relate to IREEL, which will be sued to configure GIMI services?
    8898
    8999e)  (Ilia)  what is best way to load images?  packages?
    90100
    91101f)  (Christoph)  Add iREEL (experimenter portal service); 
    92  - on 4/25:  what can this do?  can it provide MAP functions?  can it drive scripts?  (need copy to run locally) 
    93 
     102 - on 4/25:  what can this do?  can it provide MAP functions?  can it drive scripts?  (need copy to run locally)
     103 - on 5/2:  1st plan for MAP service:  base it on iREEL 2.0, installed as a public service;  where installed?  not user workspace;  not BBN;  UMassAmherst?  RENCI?  NICTA too far away.
     104 - on 5/2:  must be able to pull data from IRODS to IREEL for MAP functions.  how?
     105 - on 5/2:  if use IREEL to provide MAP functions, can 1st be used to configure the MPs/MCs, i.e., use IREEL and EC to configure the GIMI services, as if configuring a complete experiment.
     106 - on 5/2:  later, 2nd, consider extending IREEL to configure the entire experiment, i.e., interface with Flukes/OMNI, etc., to aquire resources.
     107 
     108 
    94109g)  (From GEMINI;  under discussion)  Include portal service to find other GUIs?
    95110 - on 4/25:  for 1st, not necessary 
    … …  
    169184
    170185
    171 == 7)  *** Steps towards GIMI testing ==
     186== 7)  ***Introduce IREEL service ==
     187
     188
     189(Christoph)  Add iREEL (experimenter portal service):
     190 - on 4/25:  what can this do?  can it provide MAP functions?  can it drive scripts?  (need copy to run locally)
     191 - on 5/2:  1st plan for MAP service:  base it on iREEL 2.0, installed as a public service;  where installed?  not user workspace;  not BBN;  UMassAmherst?  RENCI?  NICTA too far away.
     192 - on 5/2:  must be able to pull data from IRODS to IREEL for MAP functions.  how?
     193 - on 5/2:  if use IREEL to provide MAP functions, can 1st be used to configure the MPs/MCs, i.e., use IREEL and EC to configure the GIMI services, as if configuring a complete experiment.
     194 - on 5/2:  later, 2nd, consider extending IREEL to configure the entire experiment, i.e., interface with Flukes/OMNI, etc., to aquire resources.
     195
     196
     197
     198== 8)  +++ Steps towards GIMI testing ==
    172199
    173200a)  (Jeanne) Arrange user workspace
    … …  
    176203 - on 4/18:  plan to forward all code and plans for GEC13 to Jeanne
    177204 - on 4/18:  code leads are David and Delip;  will join future calls
     205 - on 5/2:  have sent code, scripts from GEC13 demo
     206 - on 5/2:  (Jeanne)  for tracking, have setup process on geni.net
    178207
    179208d)  (Jeanne)  What are first test scripts?  when?
    … …  
    185214
    186215
    187 == 8)  *** Steps towards GIMI tutorial at GEC14 ==
     216== 9)  +++ Steps towards GIMI tutorial at GEC14 ==
    188217
    189218a)  (Ilia) Start with GEC13 ORCA turorial?
    … …  
    197226
    198227
    199 == 9)  Use GIMI in WiMAX site  ==
     228== 10)  Use GIMI in WiMAX site  ==
    200229
    201230a)  (Mike, Cong)  Can we do something by GEC14?  what?
    202231 - on 4/25:  agree UMass Amherst WiMAX site should be up by GEC14
    203232 - on 4/25:  agree should have plan to integrate GIMI with WiMAX by GEC14, but does not need to be done
     233