Jump to: navigation, search

Difference between revisions of "Meetings/swift3"

Line 1: Line 1:
 
Meeting Agenda:  Bi-weekly on Wednesdays at 23:00 UTC  
 
Meeting Agenda:  Bi-weekly on Wednesdays at 23:00 UTC  
 +
 +
Past meeting notes: https://etherpad.openstack.org/p/swift3-meeting
  
 
Next meeting:  Aug 5, 2015
 
Next meeting:  Aug 5, 2015
Line 9: Line 11:
 
# analyze URL mapping between s3 & swift, see if s3_token middleware can provide correct account information in all cases [ntt, ibm]
 
# analyze URL mapping between s3 & swift, see if s3_token middleware can provide correct account information in all cases [ntt, ibm]
 
# look at approaches for accessing stored acl information efficiently & propose design for this [ntt, ibm]
 
# look at approaches for accessing stored acl information efficiently & propose design for this [ntt, ibm]
 
Meeting notes: https://etherpad.openstack.org/p/swift3-meeting
 

Revision as of 00:55, 23 July 2015

Meeting Agenda: Bi-weekly on Wednesdays at 23:00 UTC

Past meeting notes: https://etherpad.openstack.org/p/swift3-meeting

Next meeting: Aug 5, 2015

  • discuss compliance testing: use ceph/s3-tests or use swift3 unit and function tests
  • create first draft of publishable output (swiftstack). Will also need to define what is meant by "compatibliity", as in some case behavior is complex/nuanced. Include description of configuration used (pipeline settings, etc) to get the published results (incude swift info output?)
  • assess of state of bucket versioning vs swift versioning (swiftstack)
  • s3 acls:
  1. analyze s3-tests failures and categorize as on of (not implemented, test case failure, swift3 defect) [ibm]
  2. analyze URL mapping between s3 & swift, see if s3_token middleware can provide correct account information in all cases [ntt, ibm]
  3. look at approaches for accessing stored acl information efficiently & propose design for this [ntt, ibm]