mirror of
https://github.com/MariaDB/server.git
synced 2025-01-31 19:11:46 +01:00
0c1f97b3ab
The immediate bug was caused by a failure to recognize a correct position to stop the slave applier run in optimistic parallel mode. There were the following set of issues that the analysis unveil. 1 incorrect estimate for the event binlog position passed to is_until_satisfied 2 wait for workers to complete by the driver thread did not account non-group events that could be left unprocessed and thus to mix up the last executed binlog group's file and position: the file remained old and the position related to the new rotated file 3 incorrect 'slave reached file:pos' by the parallel slave report in the error log 4 relay log UNTIL missed out the parallel slave branch in is_until_satisfied. The patch addresses all of them to simplify logics of log change notification in either the master and relay-log until case. P.1 is addressed with passing the event into is_until_satisfied() for proper analisis by the function. P.2 is fixed by changes in handle_queued_pos_update(). P.4 required removing relay-log change notification by workers. Instead the driver thread updates the notion of the current relay-log fully itself with aid of introduced bool Relay_log_info::until_relay_log_names_defer. An extra print out of the requested until file:pos is arranged with --log-warning=3.
291 lines
7.8 KiB
Text
291 lines
7.8 KiB
Text
include/master-slave.inc
|
|
[connection master]
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
RESET MASTER;
|
|
RESET SLAVE;
|
|
connection master;
|
|
RESET MASTER;
|
|
CREATE TABLE t1 (a int primary key, b text) ENGINE=InnoDB;
|
|
INSERT INTO t1 SET a=25, b='trx0';
|
|
connection slave;
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
ALTER TABLE mysql.gtid_slave_pos ENGINE=InnoDB;
|
|
SET @old_parallel_threads=@@GLOBAL.slave_parallel_threads;
|
|
SET GLOBAL slave_parallel_threads=2;
|
|
SET @old_parallel_mode=@@GLOBAL.slave_parallel_mode;
|
|
SET GLOBAL slave_parallel_mode='optimistic';
|
|
connection slave;
|
|
SET @old_max_relay_log_size = @@global.max_relay_log_size;
|
|
SET @@global.max_relay_log_size=4096;
|
|
connection master;
|
|
BEGIN;
|
|
INSERT INTO t1 SET a=1, b='trx1';
|
|
INSERT INTO t1 SET a=2, b='trx1';
|
|
INSERT INTO t1 SET a=3, b='trx1';
|
|
INSERT INTO t1 SET a=4, b='trx1';
|
|
INSERT INTO t1 SET a=5, b='trx1';
|
|
INSERT INTO t1 SET a=6, b='trx1';
|
|
INSERT INTO t1 SET a=7, b='trx1';
|
|
INSERT INTO t1 SET a=8, b='trx1';
|
|
INSERT INTO t1 SET a=9, b='trx1';
|
|
INSERT INTO t1 SET a=10, b='trx1';
|
|
INSERT INTO t1 SET a=11, b='trx1';
|
|
INSERT INTO t1 SET a=12, b='trx1';
|
|
INSERT INTO t1 SET a=13, b='trx1';
|
|
INSERT INTO t1 SET a=14, b='trx1';
|
|
INSERT INTO t1 SET a=15, b='trx1';
|
|
INSERT INTO t1 SET a=16, b='trx1';
|
|
INSERT INTO t1 SET a=17, b='trx1';
|
|
INSERT INTO t1 SET a=18, b='trx1';
|
|
INSERT INTO t1 SET a=19, b='trx1';
|
|
INSERT INTO t1 SET a=20, b='trx1';
|
|
INSERT INTO t1 SET a=21, b='trx1';
|
|
INSERT INTO t1 SET a=22, b='trx1';
|
|
INSERT INTO t1 SET a=23, b='trx1';
|
|
INSERT INTO t1 SET a=24, b='trx1';
|
|
COMMIT;
|
|
FLUSH LOGS;
|
|
BEGIN;
|
|
UPDATE t1 SET b='trx2_0' WHERE a = 25;
|
|
UPDATE t1 SET b='trx2' WHERE a = 25;
|
|
COMMIT;
|
|
INSERT INTO t1 SET a=26,b='trx3';
|
|
*** case 1 UNTIL inside trx2
|
|
connection slave1;
|
|
BEGIN;
|
|
INSERT INTO t1 SET a= 1;
|
|
connection slave;
|
|
SELECT <pos_0> <= <pos_until> AND <pos_until> < <pos_trx2> as "pos_until < trx0 and is within trx2";
|
|
pos_until < trx0 and is within trx2
|
|
1
|
|
CHANGE MASTER TO MASTER_USE_GTID=no;
|
|
START SLAVE UNTIL MASTER_LOG_FILE = 'file_2', MASTER_LOG_POS = <pos_until>;
|
|
connection slave1;
|
|
ROLLBACK;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
SELECT count(*) = 1 as 'trx2 is committed' FROM t1 WHERE b = 'trx2';
|
|
trx2 is committed
|
|
1
|
|
SELECT count(*) = 0 as 'trx3 is not committed' FROM t1 WHERE b = 'trx3';
|
|
trx3 is not committed
|
|
1
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
*** case 2 UNTIL inside trx2
|
|
connection slave;
|
|
DELETE FROM t1 WHERE a <> 25;
|
|
UPDATE t1 SET b='trx0' WHERE a = 25;
|
|
connection slave1;
|
|
BEGIN;
|
|
INSERT INTO t1 SET a= 1;
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
SELECT <pos_0> <= <pos_until> AND <pos_until> < <pos_trx2> as "pos_until >= trx0 and is within trx2";
|
|
pos_until >= trx0 and is within trx2
|
|
1
|
|
CHANGE MASTER TO MASTER_LOG_FILE = 'file_1', MASTER_LOG_POS = <pos_trx0>, MASTER_USE_GTID=no;
|
|
START SLAVE UNTIL MASTER_LOG_FILE = 'file_2', MASTER_LOG_POS = <pos_until>;
|
|
connection slave1;
|
|
ROLLBACK;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
SELECT count(*) = 1 as 'trx2 is committed' FROM t1 WHERE b = 'trx2';
|
|
trx2 is committed
|
|
1
|
|
SELECT count(*) = 0 as 'trx3 is not committed' FROM t1 WHERE b = 'trx3';
|
|
trx3 is not committed
|
|
1
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
*** case 3 UNTIL inside trx1
|
|
connection slave;
|
|
DELETE FROM t1 WHERE a <> 25;
|
|
UPDATE t1 SET b='trx0' WHERE a = 25;
|
|
connection slave1;
|
|
BEGIN;
|
|
INSERT INTO t1 SET a= 1; # block trx1;
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
SELECT <pos_until> < <pos_0> as "pos_until before trx2 start position";
|
|
pos_until before trx2 start position
|
|
1
|
|
CHANGE MASTER TO MASTER_LOG_FILE = 'file_1', MASTER_LOG_POS = <pos_trx0>, MASTER_USE_GTID=no;
|
|
START SLAVE UNTIL MASTER_LOG_FILE = 'file_2', MASTER_LOG_POS = <pos_until>;
|
|
connection slave1;
|
|
ROLLBACK;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
SELECT count(*) = 25-1 as 'trx1 is committed' FROM t1 WHERE b = 'trx1';
|
|
trx1 is committed
|
|
1
|
|
SELECT count(*) = 0 as 'trx2 is not committed' FROM t1 WHERE b = 'trx2';
|
|
trx2 is not committed
|
|
1
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
*** case 4 Relay-log UNTIL inside trx1
|
|
connection slave;
|
|
DELETE FROM t1 WHERE a <> 25;
|
|
UPDATE t1 SET b='trx0' WHERE a = 25;
|
|
connection slave1;
|
|
BEGIN;
|
|
INSERT INTO t1 SET a= 1; # block trx1;
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
CHANGE MASTER TO MASTER_LOG_FILE = 'file_1', MASTER_LOG_POS = <pos_trx0>, MASTER_USE_GTID=no;
|
|
START SLAVE IO_THREAD;
|
|
include/wait_for_slave_io_to_start.inc
|
|
START SLAVE UNTIL RELAY_LOG_FILE = 'file_2', RELAY_LOG_POS = <pos_until>;
|
|
connection slave1;
|
|
ROLLBACK;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
SELECT count(*) = 25-1 as 'trx1 is committed' FROM t1 WHERE b = 'trx1';
|
|
trx1 is committed
|
|
1
|
|
SELECT count(*) = 0 as 'trx2 is not committed' FROM t1 WHERE b = 'trx2';
|
|
trx2 is not committed
|
|
1
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
*** case 5 Relay-log UNTIL inside a "big" trx that spawns few relay logs
|
|
connection master;
|
|
CREATE TABLE t2 (a TEXT) ENGINE=InnoDB;
|
|
FLUSH LOGS;
|
|
connection slave;
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
connection master;
|
|
BEGIN;
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
INSERT INTO t2 SET a=repeat('a',1024);
|
|
COMMIT;
|
|
INSERT INTO t2 SET a='a';
|
|
connection slave;
|
|
START SLAVE IO_THREAD;
|
|
include/wait_for_slave_io_to_start.inc
|
|
START SLAVE UNTIL RELAY_LOG_FILE = 'file_2', RELAY_LOG_POS = <pos_until>;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
include/diff_tables.inc [master:t2,slave:t2]
|
|
*** case 6 Relay-log UNTIL inside a small trx inside a sequence of relay logs
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
connection master;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
BEGIN;
|
|
DELETE FROM t2 LIMIT 1;
|
|
COMMIT;
|
|
COMMIT;
|
|
connection slave;
|
|
START SLAVE IO_THREAD;
|
|
include/wait_for_slave_io_to_start.inc
|
|
connection master;
|
|
include/sync_slave_io_with_master.inc
|
|
connection slave;
|
|
START SLAVE UNTIL RELAY_LOG_FILE = 'file_2', RELAY_LOG_POS = <pos_until>;
|
|
Proof 1: Correct stop
|
|
connection slave;
|
|
include/wait_for_slave_sql_to_stop.inc
|
|
Proof 2: Resume works out
|
|
include/start_slave.inc
|
|
connection master;
|
|
connection slave;
|
|
include/diff_tables.inc [master:t2,slave:t2]
|
|
connection slave;
|
|
include/stop_slave.inc
|
|
SET GLOBAL max_relay_log_size=@old_max_relay_log_size;
|
|
SET GLOBAL slave_parallel_mode=@old_parallel_mode;
|
|
SET GLOBAL slave_parallel_threads=@old_parallel_threads;
|
|
include/start_slave.inc
|
|
connection master;
|
|
DROP TABLE t1, t2;
|
|
connection slave;
|
|
include/rpl_end.inc
|