~drizzle-trunk/drizzle/development : changes

~drizzle-trunk/drizzle/development » Changes from revision 1405.6.1

Rev	Summary	Authors	Date
1405.6.1	This patch adds the following functionality: This patch adds the following functionality: * Removes the need to manually enable replicators in order for an applier to work. * Removes the enabled/disabled setting of both transaction applier plugins and transaction replicator plugins * Pairs a replicator with an applier into a "ReplicationStream" and removes all checks for "enabled" replicators and appliers * Allows modules that implement a TransactionApplier (such as the transaction_log module) to specify which replicator to use via a configuration variable. For instance, the transaction log module now has --transaction-log-use-replicator=[default\|filtered..] instead of the user having to do --default-replicator-enable and such * Adds a new data dictionary table for REPLICATION_STREAMS, which allows querying of activated replication-to-applier streams managed by drizzled::ReplicationServices	Jay Pipes	14 years ago
1405.4.14	merge replication-api merge replication-api	Jay Pipes	14 years ago
1405.4.13	merge trunk merge trunk	Jay Pipes	14 years ago
1405.4.12	Fix cpplint header guard. Fix cpplint header guard.	Jay Pipes	14 years ago
1405.4.11	Merge trunk Merge trunk	Jay Pipes	14 years ago
1405.4.10	OK, Sun Studio still didn't like that...seems to think that ... OK, Sun Studio still didn't like that...seems to think that inline means something different than other compilers think it is...	Jay Pipes	14 years ago
1405.4.9	Sun Studio says inlined functions must be defined before bei... Sun Studio says inlined functions must be defined before being called...	Jay Pipes	14 years ago
1405.4.8	Add Solaris atomics fixes and test files. Add replication.h ... Add Solaris atomics fixes and test files. Add replication.h header to makefile.	Jay Pipes	14 years ago
1405.4.7	* Fixes drizzled's atomics: * Fixes drizzled's atomics: - fetch_and_add() was actually add_and_fetch() - fixed to have both methods correct - compare_and_swap() was incorrect for all traits classes. Fixed to return a bool true only when the supplied value is actually swapped - fixes increment() and decrement() methods and operator+=() in outer atomics class template to call proper add_and_fetch() methods on traits classes - Now that above are fixed, removed the hacks in Query_id and TransactionLog to have query ID and the new transactoin ID start properly at 1. * Transaction messages sent over replication stream now use a real transaction ID, managed by drizzled::TransactionServices. Previously, the Query_id was being used, resulting in SELECT statements incrementing the transaction ID. * Added a test case to ensure that DDL ops are given a transaction ID and SELECT ops do not increment the transaction ID. The transaction ID will be paired with a channel ID to become the global transaction identifier. ReplicationServices will manage the pairing of channel and transaction ID and understand how far a particular subscriber node has applied.	Jay Pipes	14 years ago
1405.4.6	Add DROP TABLE to insert test to ensure test cases are idemp... Add DROP TABLE to insert test to ensure test cases are idempotent.	Jay Pipes	14 years ago
1405.4.5	Changes Query_id to start at 0, not 1, since increment() wil... Changes Query_id to start at 0, not 1, since increment() will make the first Query ID 2, not 1.	Jay Pipes	14 years ago
1405.4.4	Merge trunk Merge trunk	Jay Pipes	14 years ago
1405.4.3	merge write buffers work merge write buffers work	Jay Pipes	14 years ago
1405.4.2	Merge const removal of Session parameter in replication API. Merge const removal of Session parameter in replication API.	Jay Pipes	14 years ago
1405.4.1	Adds a vector of write buffers to the transaction log applie... Adds a vector of write buffers to the transaction log applier object. This eliminates the calls to malloc() for writing most transaction log entries (the only calls to malloc() will be when an existing write buffer is not large enough). There is a configuration setting now for the number of write buffers to create on startup. The slot algorithm is a simple modulo on the session ID.	Jay Pipes	14 years ago
1405.3.6	Here, we do two main things: Here, we do two main things: * We alter the code path inside the transaction coordinator to push transaction messages to replicators in between the prepare and commit stages of a transaction * We remove the prepare_commit_mutex in the InnoDB storage engine	Jay Pipes	14 years ago
1405.3.5	TransactionServices method names now meet code style guideli... TransactionServices method names now meet code style guidelines.	Jay Pipes	14 years ago
1405.3.4	Removes mysql_xa_recover() Removes mysql_xa_recover()	Jay Pipes	14 years ago
1405.3.3	Adds Session reference to replication API Adds Session reference to replication API	Jay Pipes	14 years ago
1405.3.2	Add replication common header file. Add replication common header file.	Jay Pipes	14 years ago

Older »