1. 12 Jun, 2013 1 commit
  2. 11 Jun, 2013 10 commits
    • Vincent Pelletier's avatar
      Use single-underscore property prefix. · 43ab06c9
      Vincent Pelletier authored
      For easier debugging, so we don't have to know how python mangles
      double-underscore properties, as introspection is reduced by __slots__
      usage.
      Suggested by Julien Muchembled <jm@nexedi.com>
      43ab06c9
    • Vincent Pelletier's avatar
      Move some work out of Message.__init__ . · fda3f093
      Vincent Pelletier authored
      So that creating an ActiveWrapper (or Method) once and reusing it to spawn
      several activities gets a larger speed-up.
      Message class API is not supposed to be used outside this module, so
      drop failing test rather than fixing it.
      fda3f093
    • Vincent Pelletier's avatar
      Declare __slots__ . · 75e7c3b4
      Vincent Pelletier authored
      Also, inherit from "object" as __slots__ are a new-style-class feature.
      75e7c3b4
    • Vincent Pelletier's avatar
      Replace __getattr__ + __[sg]etitem__ calls by a single __[sg]etattr__ call. · c6fab535
      Vincent Pelletier authored
      Also, provide one argument per line.
      Also, avoid shadowing "id" built-in.
      c6fab535
    • Vincent Pelletier's avatar
      Make registerMessage check for duplicate registration. · 2fa6bfeb
      Vincent Pelletier authored
      isMessageRegistered duplicates work done in registerMessage, so it wastes
      time when creating an activity (in the likely event the activity is not a
      duplicate).
      2fa6bfeb
    • Vincent Pelletier's avatar
      Avoid looking up portal_activities when creating a message. · 290a9b38
      Vincent Pelletier authored
      It is (likely) already known to caller, is only used to look up an option
      which is rarely enabled, and it turns out to be (relatively) expensive.
      290a9b38
    • Vincent Pelletier's avatar
      1ee39d03
    • Vincent Pelletier's avatar
      19e6cc3e
    • Vincent Pelletier's avatar
      Work around poor UPDATE use of index. · 7daaf0a5
      Vincent Pelletier authored
      UPDATE query is exected to use the existing index on (processing_node,
      priority, date) both for WHERE and ORDER BY, as is expected from
      EXPLAIN-ing the equivalent SELECT:
      
      MariaDB [erp5]> explain select uid from message_queue WHERE processing_node=0 AND date <= '2013-06-06 22:22:49' ORDER BY priority, date LIMIT 1;
      +------+-------------+---------------+------+----------------------------------------------------------+-------------------------------+---------+-------+-------+--------------------------+
      | id   | select_type | table         | type | possible_keys                                            | key                           | key_len | ref   | rows  | Extra                    |
      +------+-------------+---------------+------+----------------------------------------------------------+-------------------------------+---------+-------+-------+--------------------------+
      |    1 | SIMPLE      | message_queue | ref  | processing_node_processing,processing_node_priority_date | processing_node_priority_date | 2       | const | 26622 | Using where; Using index |
      +------+-------------+---------------+------+----------------------------------------------------------+-------------------------------+---------+-------+-------+--------------------------+
      
      If it weren't using the index for ORDER BY, "Extra" would contain
      "Using filesort".
      
      Still, UPDATE behaves differently:
      
        # User@Host: user[user] @  [10.0.0.3]
        # Thread_id: 1635880  Schema: erp5  QC_hit: No
        # Query_time: 2.668405  Lock_time: 2.460698  Rows_sent: 0  Rows_examined: 49263
        # Full_scan: No  Full_join: No  Tmp_table: No  Tmp_table_on_disk: No
        # Filesort: Yes  Filesort_on_disk: No  Merge_passes: 0
        SET TIMESTAMP=1370557446;
        UPDATE
          message_queue
        SET
          processing_node=12
        WHERE
          processing_node=0
          AND DATE <= '2013-06-06 22:24:04'
          ORDER BY
          priority, DATE
        LIMIT 1;
      
      So change the UPDATE..SELECT pattern into a SELECT FOR UPDATE..UPDATE
      pattern, so SELECT's correct execution plan is used.
      7daaf0a5
    • Kazuhiko Shiozaki's avatar
      explicitly specify update_catalog argument in business template installation. · 3b74878d
      Kazuhiko Shiozaki authored
      that should be part of commit 9b5c74fb ('force' argument in
      BusinessTemplate._install() should not mean clear catalog. we can and
      should use 'update_catalog' argument instead to clear catalog).
      3b74878d
  3. 10 Jun, 2013 6 commits
  4. 07 Jun, 2013 5 commits
  5. 06 Jun, 2013 4 commits
  6. 05 Jun, 2013 2 commits
    • Kazuhiko Shiozaki's avatar
      respect user's input for source or destionation in Base_addEvent. · 61e73e83
      Kazuhiko Shiozaki authored
      before, user's input is completely ignored and calculated good values
      are automatically set. and the commit fc54fbaa somehow tried to
      respect user's source input, but it caused worse result for incoming
      case. so this commit reverts fc54fbaa and set source and destination
      just as user's input if not empty.
      61e73e83
    • Arnaud Fontaine's avatar
      Avoid regeneration of classes when starting a node within a cluster with already started nodes. · 20362ede
      Arnaud Fontaine authored
      When starting a node, ERP5Site.__of__ calls ComponentTool.reset(), which was
      never synchronised so it returned True (without generating a new cookie at its
      level) and causing synchronizeDynamicModules to be called with
      force=True. This generated a new cookie and leading to dynamic classes being
      meaninglessly regenerated on all nodes.
      
      Instead, modify ComponentTool.reset() behavior so it *always* reset Portal
      Type classes when Components are reset (as it should have always been as the
      class inheritance may have been modified) and force regeneration of Portal
      Type classes in this method if reset is True.
      Signed-off-by: Vincent Pelletier's avatarVincent Pelletier <vincent@nexedi.com>
      20362ede
  7. 04 Jun, 2013 6 commits
    • Julien Muchembled's avatar
    • Vincent Pelletier's avatar
      Drop unneeded criterion. · 0d2e4dd6
      Vincent Pelletier authored
      0d2e4dd6
    • Vincent Pelletier's avatar
      Use REPLACE rather than INSERT for DELETE..INSERT indexation pattern. · 5d3ea021
      Vincent Pelletier authored
      When a document gets rows inserted in category table while there was none
      before (typically: first document indexation), it may trigger
        IntegrityError: (1062, "Duplicate entry '...' for key 'PRIMARY'")
      because in the DELETE..INSERT pattern, DELETE finds no matching rows and
      does not gap-lock (because we enable innodb_locks_unsafe_for_binlog), then
      the second INSERT happening, chronologically speaking, waits for the
      transaction of the first to e committed, and on commit it causes such
      duplicate key error.
      
      A transient visible effect of this change is that if both competing
      indexations see a different document state (because document got modified
      in some 3rd transaction), the union of the resulting set of rows will be
      visible until the reindexation which must have been triggered by the 3rd
      transaction gets executed, at which point only the latest set will be
      visible.
      
      A similar issue exists before this change, with stricter conditions: it
      needs the intersection of both sets to be empty, because a non-empty
      intersection causes the duplicate key error solved here.
      
      This change has been measured to improve scalability of an invoice building
      test case (naturally triggering indexations) starting from ~12 activity
      nodes:
         8 nodes:  +1.4% invoices/hour
        12 nodes:  +9.5% invoices/hour
        16 nodes: +12.3% invoices/hour
      (values are the difference between DELETE..INSERT and DELETE..REPLACE)
      5d3ea021
    • Julien Muchembled's avatar
      ZSQLCatalog: fix identation in EntireQuery · 4ec66892
      Julien Muchembled authored
      asSQLExpression() failed when executed several times on the same instance,
      because too much code was skipped when 'column_map' is cached.
      4ec66892
    • Vincent Pelletier's avatar
      Setter argument is required. · a5535be1
      Vincent Pelletier authored
      a5535be1
    • Julien Muchembled's avatar
      0eb3c1e8
  8. 03 Jun, 2013 1 commit
  9. 31 May, 2013 1 commit
  10. 30 May, 2013 1 commit
  11. 29 May, 2013 3 commits