got LOAD DATA INFILE replication from 3.23 master to 4.0 slave to work
for some reason query_cache now fails - will push anyway and document the
failure if I can repeat it.
fixed bug in LOAD DATA FROM MASTER
fixed rpl000001.result
Slave now replicates 3.23 master, with the exception of LOAD DATA INFILE,
which is still buggy. Will push this one after the pull/merge
I did not know I can't do chmod ...
So, it killed some of my changes.
But I am smarter, and as I have first saved my work, before BK ruined
it, so here it comes ...
fixed apparent typo in opt_sum.cc
I will push this Changeset although it fails select test - failing select
is better than the code that does not compile.
Bug fix when using --no-deaults
Sets ref_length to right value (faster rnd_pos() handling in InnoDB).
Fixed problem with multi-table-delete and InnoDB
Fixed problem with truncate table, which required a COMMIT before.
Fixed multi-table-delete bug with InnoDB.
Remove not used index from EXPLAIN
Assume selectivity of any key prefix is better than 1 / 100, regardless of what the estimator returns; this balances the pessimistic assumption of the optimizer thatany secondary key access to the full row incurs a page read from disk
Test of unsigned BIGINT values
Fixes for queries-per-hour
Cleanup of replication code (comments and portability fixes)
Make most of the binary log code 4G clean
Changed syntax for GRANT ... QUERIES PER HOUR
fixed a couple of bugs with SEQ_READ_APPEND cache
rpl000016 still has non-deterministic result, but I am going to commit and
push since what I have is now better than what is in the main repository
Faster 'read_first_row' (Fixes slow 'preparing' state)
Read constant tables earlier, which provides better optimzations when using tables with <=1 row.
This also fixes a complicated bug involving const tables.
fixed too quick timeout in mysql-test-run which caused a race with the
new server getting started before the old one completely finished
shutdown. This should fix the pid warning we've been getting as well as
inconsistent results when running tests with the manager
first pull, merge,test, and get it to work.
The main change is the new replication code - now we have two slave threads
SQL thread and I/O thread. I have also re-written a lot of the code to
prepare for multi-master implementation.
I also documented IO_CACHE quite extensively and to some extend, THD class.
