1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
|
###############################################################################
# Bug#22247668 SLAVE IS ~10X SLOWER TO EXECUTE SET OF STATEMENTS COMPARED TO
# MASTER RBR
# Problem: A new style of locking is implemented in Innodb. Starting from 5.6,
# Innodb uses this new style of locking for all the cases except for the case
# where for a simple (single/multi) row INSERTs, it fall back to old style
# locking if another transaction has already acquired the AUTOINC lock on behalf of
# a LOAD FILE or INSERT...SELECT etc. type of statement. But on
# Slave, in RBR format, it is always using old style auto inc
# algorithm.
#
# Steps to reproduce:
# 1) Setup DEBUG simulation point on Slave to bring the server down
# if the INSERT enters old style autoinc locking method.
#
# 2) Execute AUTOINC related work on Master.
#
# 3) Make sure that sync on slave happens without any issues.
#
###############################################################################
--source include/have_debug.inc
--source include/have_binlog_format_row.inc
--source include/master-slave.inc
# Initial setup
CREATE TABLE t (i INT AUTO_INCREMENT PRIMARY KEY) ENGINE=InnoDB;
--source include/sync_slave_sql_with_master.inc
# Step-1 : Setup DEBUG simulation point on Slave to bring the server down
# if the INSERT enters old style autoinc locking method.
SET @save_debug=@@debug;
SET GLOBAL DEBUG='+d,die_if_autoinc_old_lock_style_used';
# Step-2 :Execute AUTOINC related work on Master.
--source include/rpl_connection_master.inc
INSERT INTO t VALUES (1);
DROP TABLE t;
# Step-3: Due to above DEBUG simulation point, server will go down if it enters
# old autoinc lock style. After fix, sync on Slave happens without any issues.
--source include/sync_slave_sql_with_master.inc
# Cleanup
# Reset the simulation point on Slave.
SET GLOBAL DEBUG=@save_debug;
--source include/rpl_end.inc
|
