Difference between revisions of "Meeting Minutes 2014-04-16"

From OpenCompute
Jump to: navigation, search
(Created page with "Meeting Time : 11AM-12PM EDT on 4/16/2014 - Rajeev Agrawala Agenda: - Final Pass on Hardware Management Spec - Optional features to be included in spec - OCP ID spec ...")
 
 
Line 3: Line 3:
 
Agenda:
 
Agenda:
 
      
 
      
- Final Pass on Hardware Management Spec
+
* Final Pass on Hardware Management Spec
- Optional features to be included in spec
+
* Optional features to be included in spec
- OCP ID spec discussion - Hank Bruning
+
* OCP ID spec discussion - Hank Bruning
 
   
 
   
- Please add more topics above
 
 
   
 
   
 
 
      
 
      
 
New Items  (If you add a new item please add the date you added it along with your name)
 
New Items  (If you add a new item please add the date you added it along with your name)

Latest revision as of 17:04, 17 April 2014

Meeting Time : 11AM-12PM EDT on 4/16/2014 - Rajeev Agrawala

Agenda:

  • Final Pass on Hardware Management Spec
  • Optional features to be included in spec
  • OCP ID spec discussion - Hank Bruning


New Items (If you add a new item please add the date you added it along with your name)

Attendees:

   Hank Bruning(JBlade)
   SuiLun Lam (NEC)
   Grant Richard (GS)
   Earl Medeiros (ZT)
   Nathan Rockhold (NodePrime)
   Paul Vancil (Dell)
   Rajeev Agrawala (GS)
   Yuval Itkin (Mellanox)
   Per Brashers (Yttibrium, and OCP Storage Chair)
   Hari Ramachandran(AMD)
   Paul Gleichauf (Lituus)
   Tom Slaight (Intel)

Please remind everyone about the Upcoming events and event calendar - http://www.opencompute.org/community/events/ Amber Graner - 2/19/14, 3/17/14

Minutes:

  • Grant walked through the final draft. Please send in your changes and then we will publish this spec
  • Optional Features
    • Shared BMC -- what is the spec for mezz card or dedicated management network
    • Graphical console -- VGA? only, full res?, utilizing low bandwidth so need to be careful. BMC..
    • Inventory Expanded
  2nd the opinion that peripherals should be enumerated -Per
  • Identifying revisions and type. Type proposal was a 4 byte number with first byte identifying broader category like motherboard, networking or storage etc and remaining 3 bytes pointing to a spec submission to OCP in that broad category. Remaining information like vendor and device etc can be obtained by the spec. Final agreement was to use a 2 byte number instead of 3 due to byte alignment reasons.