Configuration Reference

rubyrep comes with sensible defaults for practically everything but the database connection parameters. Nevertheless a wide range of behaviors can be configured. This page lists all available configuration settings.

Global vs. table specific settings

The configuration parameters come in two types

  • global (apply to all tables)
  • table specific

Global configuration settings are specified as in the following example:

RR::Initializer::run do |config|
  ...
  config.options[:sync_conflict_handling] = :right_wins
  ...
end

Table specific options can be specified either as global default settings (same form as above) or for specific tables:

...  
config.add_table_options 'emp', :sync_conflict_handling => :left_wins
config.include_tables 'dept', 
  :sync_conflict_handling => :left_wins, 
  :logged_sync_events => :ignored_changes
...   

In the first form above, only the option is set; in the second the “dept” table will also be added to the configured tables list.

List of configuration settings

Setting Table Specific Default value Description
General Settings      
:proxy_block_size yes 1000 For proxied scans / syncs: number of rows that are checksummed
:row_buffer_size yes 1000 Number of rows that are ready into memory at a single time. Should be smaller for tables with very large columns.
:commit_frequency   1000 After how many changes the open transaction is committed and a new one started.
:use_ansi yes yes[1] If the scan / sync progress bars use ANSI codes for a slicker display.
:key yes   Manually specify the unique columns of a table. More information here.
:auto_key_limit yes 0 Convenience option. Maximum number of columns of a table without primary key up to which the combination of all columns is considered unique. More information here.
:event_filter yes   Filter which differences to sync / replicate. Implement custom sync / replicate actions. More information here.
Sync specific settings      
:table_ordering   true If the sync sequence of tables is modified to reduce foreign key conflicts.
:before_table_sync yes   Custom behavior that is executed before a table sync. More information here.
:after_table_sync yes   Same as :before_table_sync (but executed after the table sync).
Replication specific settings      
:rep_prefix   ‘rr’ The prefix that is put in front of all database objects created by rubyrep.
:key_sep   ‘|’ The string that separates columns in the key column of the change log table. Must not occure in the key columns of the replicated tables. Can consist of multiple characters.
:initial_sync yes true If true, syncs the tables when starting replication.
Disable with care! (I. e. only if sure that tables have the same data in both databases before starting replication
:adjust_sequences yes true If true, adjusts the sequences of the tables to avoid duplicate key problems.
:sequence_increment yes 2 The increment of adjusted sequences.
:left_sequence_offset yes 0 Ensures that adjusted sequences in the “left” database only generate numbers with number % sequence_increment == left_sequence_offset
:right_sequence_offset yes 1 Same as :left_sequence_offset (but for “right” database).
:replication_interval   1 Time in seconds between replication runs. (Can also be parts of a second.)
Note: A replication run without any unreplicated changes is very cheap.

[1] True unless output is not going to a terminal OR running under Windows.

Note There are a few more settings that have not been mentioned here. These require a good understanding of rubyrep interna. They are documented in the RDoc.