blob: 7256fed656fc6bbdab86d1b7d0c0d5892af8be85 (
plain)
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
|
# Verify that after CHANGE MASTER, replication (I/O thread and SQL
# thread) restart from where SQL thread left, not from where
# I/O thread left (some old bug fixed in 4.0.17)
source include/master-slave.inc;
connection master;
# Make SQL slave thread advance a bit
create table t1(n int);
sync_slave_with_master;
select * from t1;
# Now stop it and make I/O slave thread be ahead
stop slave sql_thread;
connection master;
insert into t1 values(1);
insert into t1 values(2);
save_master_pos;
connection slave;
--real_sleep 3 # wait for I/O thread to have read updates
stop slave;
--replace_result $MASTER_MYPORT MASTER_PORT
--replace_column 1 # 7 # 8 # 9 # 16 # 23 # 33 # 35 # 36 #
query_vertical SHOW SLAVE STATUS;
change master to master_user='root';
--replace_result $MASTER_MYPORT MASTER_PORT
--replace_column 1 # 7 # 8 # 9 # 16 # 23 # 33 # 35 # 36 #
query_vertical SHOW SLAVE STATUS;
start slave;
sync_with_master;
select * from t1;
connection master;
drop table t1;
save_master_pos;
connection slave;
sync_with_master;
# End of 4.1 tests
|
