Changes between Initial Version and Version 1 of june061810agenda


Ignore:
Timestamp:
06/18/10 16:29:02 (10 years ago)
Author:
hmussman@bbn.com
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • june061810agenda

    v1 v1  
     1Cluster D conference call on Friday, June 18, 2010
     2
     3
     4Attendees:   [[BR]]
     5
     6Ilia Baldine  (RENCI)[[BR]]
     7Anirban Mandal [[BR]]
     8Chris Heermann[[BR]]
     9
     10Harry Mussman  (GPO)  [[BR]]
     11
     12
     13
     14Mike Zink (UMass) [[BR]]
     15Emmanuel Cecchet [[BR]]
     16Brian Lynn[[BR]]
     17David Irwin[[BR]]
     18
     19Denise Gurkan  (Univ Houston)
     20
     21Michael Wang  (Columbia) [[BR]]
     22
     23Onjin  (Oklahoma)  [[BR]]
     24Wenjie[[BR]]
     25
     26Wenjie (Ohio State)  [[BR]]
     27
     28Hongwei (Wayne State)[[BR]]
     29
     30Rudra Dutta  (NCState) [[BR]]
     31
     32Agenda:  [[BR]]
     33
     341) Taking stock of known unresolved problems:[[BR]]
     35+  Brian Lynn:  Still has issues with pre-mature close[[BR]]
     36+  Ilia:  Study more.[[BR]]
     37+  David:  Eucalyptus cluster at UMass;  need to connect via new clearinghouse, before GEC8
     38+  Ilia:  Have clearinghouse up by next week.
     39+  David:  Needs to check new broker, explain issues with policy and make request.
     40
     412) New networking driver structure[[BR]]
     42+  Denise:  Shade working on driver for 3400
     43+  Ilia:  New document on drivers, quite detailed.
     44
     453) GEC8 demos  [[BR]]
     46+  DICLOUD:  yes;  ticket
     47+  ViSE:  yes;  no ticket
     48+  DOME:  no
     49+  IMF:  yes
     50+  ORCA/BEN:
     51-  OpenFlow demo RENCI and GIST Korea
     52-  Programmable router, on multiple site hosts,  practice for GEC9
     53+  Kansei:  yes;  un cross-site demo, sensor nodes and latency through
     54+  OKGEM:  yes, possibly controllable robot
     55+  LEARN:  Integration of 3400 into LEARN, at Univ Houston and Rice
     56
     574)  GENI Solictation 3:  [[BR]]
     58+  Ilia:  Considering GENI racks
     59-  Many Cluster D sites
     60-  Each site with separate network connectivity 
     61-  Spaces in lab or university data center? 
     62-  Connectivity via many (10+) VLANs? 
     63-  Need good connectivity into sites   
     64+  David:  Space, power, cooling requirements for each rack?
     65-  Ilia:  19" full height rack
     66-  GENI enabled open-flow switches
     67+  Ilia:  Purchased by project, sent out to each site
     68-  Agreement with manufacturer, who sends it out.
     69-  Harry:  See current WiMAX mesoscale project
     70+  Ilia:  Two CFs:  ORCA and protoGENI  Can be accessed by either
     71-  Local interface, but also managed by central site (e.g., RENCI)
     72+  Ilia:  How many sites?  40 racks too many for level of funding
     73-  Perhaps 15 sites, each with one or two racks
     74-  Need sites on the west coast
     75+  Ilia:  Connectivity to backbone costs
     76-  Campus?   
     77-  Regional networks?   
     78-  Perhaps:  FrameNet wherever possible, otherwise Starlight
     79
     80
     81Last call A) Connectivity to Wayne State and Ohio State[[BR]]
     82+  Joe:  Ready to go from Starlight to Merit racks; [[BR]]
     83+  Hongwei:  Merit to OARnet to Ohio State;  also Merit to Wayne State[[BR]]
     84+ Ilia:  Working with MAX on interoperability[[BR]]
     85+ Ilia:  working toward translation from protoGENI to NDL  [[BR]]
     86- Harry: Srikanth also working on topic[[BR]]
     87
     88Last call B)  GEC9 demo proposals, accepted: [[BR]]
     89a)  Joe (Northwestern):  iGENI, HP and Ericsson:  video transcoding on clouds over dynamically provisioned networks.  Prototype at GEC8.  HP wants to include OpenFlow. [[BR]]
     90b)  David (UMass):  Now-casting (short-term weather forcasting), GEC7 demo plus more computation [[BR]]
     91c)  Jeff Chase (Duke):  Dynamically stand-up ASs with custom click routers      [[BR]]