1. 01 Oct, 2010 7 commits
  2. 30 Sep, 2010 4 commits
  3. 24 Sep, 2010 3 commits
  4. 23 Sep, 2010 6 commits
  5. 16 Sep, 2010 3 commits
  6. 07 Sep, 2010 1 commit
  7. 05 Sep, 2010 9 commits
  8. 03 Sep, 2010 5 commits
  9. 02 Sep, 2010 2 commits
    • Vincent Pelletier's avatar
      Make Storage.load generate a network barrier to master node. · f2720778
      Vincent Pelletier authored
      This makes sure all invalidations about loaded object up to its actual load
      have been received by client, and forwarded to ZODB.
      This fixes a potential problem of a transaction being able to see objects
      committed by another transaction after it started itself.
      
      git-svn-id: https://svn.erp5.org/repos/neo/trunk@2286 71dcc9de-d417-0410-9af5-da40c76e7ee4
      f2720778
    • Vincent Pelletier's avatar
      Make undo implementation work with replication. · 7835c09e
      Vincent Pelletier authored
      The problem with previous implementation was that each storage locally
      decided what undo actually did to data. This causes problems when a
      storage doesn't have a complete view of past transaction but accepts write
      queries, ie when it replicates.
      This implementation reduces the decision to a readable subset of storage
      nodes (which are hence not replicating), and then sends that decision to
      all storage nodes, hence fixing the issue.
      Also, DatabaseManager.storeTransaction now consistently expects object's
      value_serial to be packed, not an integer.
      
      git-svn-id: https://svn.erp5.org/repos/neo/trunk@2285 71dcc9de-d417-0410-9af5-da40c76e7ee4
      7835c09e