MDEV-32168: slave_error_param condition is never checked from the wai…

…t_for_slave_param.inc Fix some random test failures following MDEV-32168 push. Don't blindly set $rpl_only_running_threads in many places. Instead explicit stop only the IO or SQL thread, as appropriate. Setting it interfered with rpl_end.inc in some cases. Rather than clearing it afterwards, better to not set it at all when it is not needed, removing ambiguity in the test about the state of the replication threads. Don't fail the test if include/stop_slave_io.inc finds an error in the IO thread after stop. Such errors can be simply because slave stop happened in the middle of the IO thread's initial communication with the master. Signed-off-by: Kristian Nielsen <knielsen@knielsen-hq.org>
MariaDB · Nov 28, 2023 · ea4bcb9 · ea4bcb9
1 parent 1ffa8c5
commit ea4bcb9
Show file tree

Hide file tree

Showing 37 changed files with 66 additions and 72 deletions.
diff --git a/mysql-test/include/wait_for_slave_io_to_stop.inc b/mysql-test/include/wait_for_slave_io_to_stop.inc
@@ -28,6 +28,9 @@
 #     there is an error in the IO thread.
 #     (If an error is _always_ expected, a better alternative might be to
 #     use wait_for_slave_io_error.inc instead of this file).
+#     Note: This is currently always enabled, since a simple STOP SLAVE
+#     IO_THREAD can cause an error if it interrupts the slave's initial
+#     communication with the master (MDEV-32892).
 #
 #   $rpl_debug
 #     See include/rpl_init.inc
@@ -39,12 +42,15 @@
 
 --let $slave_param= Slave_IO_Running
 --let $slave_param_value= No
-if (!$rpl_allow_error)
-{
-  --let $slave_error_param= Last_IO_Errno
-}
+--let $_io_stop_save_allow_error= $slave_error_param
+# Disabled, as IO errors are left behind when a normal STOP SLAVE interrupts
+# the initial communication between the IO thread and the master (MDEV-32892).
+#if (!$rpl_allow_error)
+#{
+#  --let $slave_error_param= Last_IO_Errno
+#}
 --source include/wait_for_slave_param.inc
---let $slave_error_param=
+--let $slave_error_param= $_io_stop_save_allow_error
 
 
 --let $include_filename= wait_for_slave_io_to_stop.inc

diff --git a/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_coords.result b/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_coords.result
@@ -63,7 +63,7 @@ include/wait_for_slave_io_error.inc [errno=1236]
 SHOW TABLES;
 Tables_in_test
 table1_no_encryption
-include/stop_slave.inc
+include/stop_slave_sql.inc
 reset slave;
 ##########
 # Cleanup

diff --git a/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_coords.test b/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_coords.test
@@ -137,8 +137,7 @@ SHOW TABLES;
 
 --disable_connect_log
 # IO thread is stopped, stop SQL thread only
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 --enable_connect_log
 reset slave;
 

diff --git a/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_gtid.result b/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_gtid.result
@@ -62,7 +62,7 @@ include/wait_for_slave_io_error.inc [errno=1236]
 # ..success
 SHOW TABLES;
 Tables_in_test
-include/stop_slave.inc
+include/stop_slave_sql.inc
 reset slave;
 ##########
 # Cleanup

diff --git a/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_gtid.test b/mysql-test/suite/binlog_encryption/encrypted_master_switch_to_unencrypted_gtid.test
@@ -133,8 +133,7 @@ SHOW TABLES;
 
 --disable_connect_log
 # IO thread is stopped, wait for SQL thread to be stopped
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 --enable_connect_log
 reset slave;
 

diff --git a/mysql-test/suite/binlog_encryption/rpl_mixed_binlog_max_cache_size.result b/mysql-test/suite/binlog_encryption/rpl_mixed_binlog_max_cache_size.result
@@ -192,7 +192,7 @@ SET GLOBAL max_binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL max_binlog_stmt_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_stmt_cache_size= ORIGINAL_VALUE;
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/start_slave.inc
 connection master;
 connection slave;

diff --git a/mysql-test/suite/binlog_encryption/rpl_parallel_ignored_errors.result b/mysql-test/suite/binlog_encryption/rpl_parallel_ignored_errors.result
@@ -36,7 +36,7 @@ connection con_temp2;
 COMMIT;
 connection server_2;
 include/wait_for_slave_sql_error.inc [errno=1062]
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/assert.inc [table t1 should have zero rows where a>32]
 SELECT * FROM t1 WHERE a>32;
 a

diff --git a/mysql-test/suite/multi_source/gtid_slave_pos.result b/mysql-test/suite/multi_source/gtid_slave_pos.result
@@ -99,7 +99,7 @@ include/wait_for_slave_to_stop.inc
 set default_master_connection = 'slave2';
 include/wait_for_slave_sql_error.inc [errno=1942]
 STOP SLAVE;
-include/stop_slave.inc
+include/wait_for_slave_io_to_stop.inc
 set default_master_connection = 'slave1';
 START SLAVE;
 include/wait_for_slave_to_start.inc

diff --git a/mysql-test/suite/multi_source/gtid_slave_pos.test b/mysql-test/suite/multi_source/gtid_slave_pos.test
@@ -121,8 +121,7 @@ set default_master_connection = 'slave2';
 --let $slave_sql_errno= 1942
 --source include/wait_for_slave_sql_error.inc
 STOP SLAVE;
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/wait_for_slave_io_to_stop.inc
 set default_master_connection = 'slave1';
 START SLAVE;
 --source include/wait_for_slave_to_start.inc

diff --git a/mysql-test/suite/rpl/include/mdev-31448_conservative.inc b/mysql-test/suite/rpl/include/mdev-31448_conservative.inc
@@ -36,7 +36,8 @@ insert into t1 values (3);
 --connection slave
 --source include/start_slave.inc
 
---let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(-1)' and  command LIKE 'Slave_worker';
+# Wildcard for `state` as it depends on whether WSREP is compiled in or not.
+--let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(%)' and  command LIKE 'Slave_worker';
 --source include/wait_condition.inc
 --let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Waiting for prior transaction to commit%' and  command LIKE 'Slave_worker';
 --source include/wait_condition.inc

diff --git a/mysql-test/suite/rpl/include/mdev-31448_optimistic.inc b/mysql-test/suite/rpl/include/mdev-31448_optimistic.inc
@@ -55,7 +55,8 @@ drop table t2;
 --source include/start_slave.inc
 
 --echo # wait for T1
---let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(-1)' and  command LIKE 'Slave_worker';
+# Wildcard for `state` as it depends on whether WSREP is compiled in or not.
+--let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(%)' and  command LIKE 'Slave_worker';
 --source include/wait_condition.inc
 
 --echo # wait for T2

diff --git a/mysql-test/suite/rpl/include/rpl_binlog_max_cache_size.test b/mysql-test/suite/rpl/include/rpl_binlog_max_cache_size.test
@@ -433,8 +433,7 @@ source include/show_binlog_events.inc;
 --eval SET GLOBAL binlog_stmt_cache_size= $old_binlog_stmt_cache_size
 
 # SQL slave is stopped, stop only IO thread
---let $rpl_only_running_threads= 1
-source include/stop_slave.inc;
+source include/stop_slave_io.inc;
 source include/start_slave.inc;
 
 connection master;

diff --git a/mysql-test/suite/rpl/include/rpl_start_stop_slave.test b/mysql-test/suite/rpl/include/rpl_start_stop_slave.test
@@ -207,8 +207,7 @@ START SLAVE;
 --let $status_items= Last_IO_Errno, Last_IO_Error
 --source include/show_slave_status.inc
 
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 RESET SLAVE;
 
 --connection master

diff --git a/mysql-test/suite/rpl/r/mdev-31448_kill_ooo_finish_optimistic.result b/mysql-test/suite/rpl/r/mdev-31448_kill_ooo_finish_optimistic.result
@@ -42,7 +42,8 @@ include/save_master_gtid.inc
 connection slave;
 #
 # Cleanup
-include/stop_slave.inc
+include/wait_for_slave_sql_to_stop.inc
+include/stop_slave_io.inc
 set @@global.slave_parallel_threads= 0;
 set @@global.slave_parallel_mode= conservative;
 set @@global.innodb_lock_wait_timeout= 50;

diff --git a/mysql-test/suite/rpl/r/rpl_connection.result b/mysql-test/suite/rpl/r/rpl_connection.result
@@ -6,7 +6,7 @@ include/stop_slave.inc
 CHANGE MASTER TO MASTER_USER= '', MASTER_PASSWORD= '';
 START SLAVE;
 include/wait_for_slave_io_error.inc [errno=1045, 1593]
-include/stop_slave.inc
+include/stop_slave_sql.inc
 CHANGE MASTER TO MASTER_USER= 'root', MASTER_PASSWORD= '';
 START SLAVE;
 include/rpl_end.inc
diff --git a/mysql-test/suite/rpl/r/rpl_domain_id_filter_io_crash.result b/mysql-test/suite/rpl/r/rpl_domain_id_filter_io_crash.result
@@ -135,7 +135,7 @@ i
 2
 3
 SET @@global.debug_dbug=@saved_dbug;
-include/stop_slave.inc
+include/stop_slave_sql.inc
 DO_DOMAIN_IDS (BEFORE)     : 
 IGNORE_DOMAIN_IDS (BEFORE) : 
 CHANGE MASTER TO IGNORE_DOMAIN_IDS=(1), MASTER_USE_GTID=slave_pos;
@@ -204,7 +204,7 @@ i
 10
 11
 SET @@global.debug_dbug=@saved_dbug;
-include/stop_slave.inc
+include/stop_slave_sql.inc
 DO_DOMAIN_IDS (BEFORE)     : 
 IGNORE_DOMAIN_IDS (BEFORE) : 1
 CHANGE MASTER TO IGNORE_DOMAIN_IDS=(), MASTER_USE_GTID=slave_pos;
@@ -287,7 +287,7 @@ i
 16
 17
 SET @@global.debug_dbug=@saved_dbug;
-include/stop_slave.inc
+include/stop_slave_sql.inc
 DO_DOMAIN_IDS (BEFORE)     : 
 IGNORE_DOMAIN_IDS (BEFORE) : 1
 CHANGE MASTER TO IGNORE_DOMAIN_IDS=(), MASTER_USE_GTID=slave_pos;
@@ -384,7 +384,7 @@ i
 22
 23
 SET @@global.debug_dbug=@saved_dbug;
-include/stop_slave.inc
+include/stop_slave_sql.inc
 DO_DOMAIN_IDS (BEFORE)     : 
 IGNORE_DOMAIN_IDS (BEFORE) : 
 CHANGE MASTER TO IGNORE_DOMAIN_IDS=(1), MASTER_USE_GTID=slave_pos;

diff --git a/mysql-test/suite/rpl/r/rpl_gtid_errorlog.result b/mysql-test/suite/rpl/r/rpl_gtid_errorlog.result
@@ -23,7 +23,8 @@ INSERT INTO t1 VALUES (2);
 SET sql_log_bin=1;
 START SLAVE;
 include/wait_for_slave_sql_error.inc [errno=1062]
-include/stop_slave.inc
+include/wait_for_slave_io_to_start.inc
+include/stop_slave_io.inc
 SET GLOBAL gtid_slave_pos= "0-1-100";
 include/start_slave.inc
 SELECT * FROM t1 ORDER BY a;
@@ -39,7 +40,7 @@ REPLACE INTO t1 VALUES (5);
 SET debug_dbug= @dbug_save;
 connection slave;
 include/wait_for_slave_sql_error.inc [errno=1590]
-include/stop_slave.inc
+include/stop_slave_io.inc
 SET sql_slave_skip_counter=1;
 include/start_slave.inc
 SELECT * FROM t1 ORDER BY a;

diff --git a/mysql-test/suite/rpl/r/rpl_gtid_startpos.result b/mysql-test/suite/rpl/r/rpl_gtid_startpos.result
@@ -36,7 +36,7 @@ CHANGE MASTER TO master_host = '127.0.0.1', master_port = MASTER_PORT,
 MASTER_USE_GTID=CURRENT_POS;
 START SLAVE;
 include/wait_for_slave_io_error.inc [errno=1236]
-include/stop_slave.inc
+include/stop_slave_sql.inc
 CHANGE MASTER TO master_host = '127.0.0.1', master_port = MASTER_PORT,
 MASTER_LOG_FILE="master-bin.000003", MASTER_LOG_POS=4;
 include/start_slave.inc

diff --git a/mysql-test/suite/rpl/r/rpl_heartbeat_basic.result b/mysql-test/suite/rpl/r/rpl_heartbeat_basic.result
@@ -226,7 +226,7 @@ connection slave;
 call mtr.add_suppression("Slave SQL.*Duplicate entry .1. for key .PRIMARY.. on query.* error.* 1062");
 call mtr.add_suppression("Slave SQL.*Request to stop slave SQL Thread received while applying a group that has non-transactional changes; waiting for completion of the group");
 Heartbeat events are received while sql thread stopped (1 means 'yes'): 1
-include/stop_slave.inc
+include/stop_slave_io.inc
 DROP TABLE t1;
 
 *** Master send to slave ***

diff --git a/mysql-test/suite/rpl/r/rpl_mdev_17614.result b/mysql-test/suite/rpl/r/rpl_mdev_17614.result
@@ -29,8 +29,7 @@ SELECT * FROM t1;
 a	b	c
 1	1	1
 2	2	3
-stop slave;
-include/stop_slave.inc
+include/stop_slave_io.inc
 reset slave;
 connection master;
 reset master;
@@ -189,8 +188,7 @@ SELECT * FROM t1;
 a	b	c
 1	1	1
 2	2	3
-stop slave;
-include/stop_slave.inc
+include/stop_slave_io.inc
 reset slave;
 connection master;
 reset master;

diff --git a/mysql-test/suite/rpl/r/rpl_mixed_binlog_max_cache_size.result b/mysql-test/suite/rpl/r/rpl_mixed_binlog_max_cache_size.result
@@ -192,7 +192,7 @@ SET GLOBAL max_binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL max_binlog_stmt_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_stmt_cache_size= ORIGINAL_VALUE;
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/start_slave.inc
 connection master;
 connection slave;

diff --git a/mysql-test/suite/rpl/r/rpl_parallel_ignored_errors.result b/mysql-test/suite/rpl/r/rpl_parallel_ignored_errors.result
@@ -36,7 +36,7 @@ connection con_temp2;
 COMMIT;
 connection server_2;
 include/wait_for_slave_sql_error.inc [errno=1062]
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/assert.inc [table t1 should have zero rows where a>32]
 SELECT * FROM t1 WHERE a>32;
 a

diff --git a/mysql-test/suite/rpl/r/rpl_row_binlog_max_cache_size.result b/mysql-test/suite/rpl/r/rpl_row_binlog_max_cache_size.result
@@ -191,7 +191,7 @@ SET GLOBAL max_binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL max_binlog_stmt_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_stmt_cache_size= ORIGINAL_VALUE;
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/start_slave.inc
 connection master;
 connection slave;

diff --git a/mysql-test/suite/rpl/r/rpl_row_corruption.result b/mysql-test/suite/rpl/r/rpl_row_corruption.result
@@ -14,7 +14,7 @@ connection master;
 UPDATE t1_11753004, t2_11753004 SET t1_11753004.c1=3, t2_11753004.c1=4 WHERE t1_11753004.c1=1 OR t2_11753004.c1=2;
 connection slave;
 include/wait_for_slave_sql_error.inc [errno=1593 ]
-include/stop_slave.inc
+include/stop_slave_io.inc
 SET @@global.debug_dbug=@saved_debug;
 include/start_slave.inc
 connection master;

diff --git a/mysql-test/suite/rpl/r/rpl_stm_binlog_max_cache_size.result b/mysql-test/suite/rpl/r/rpl_stm_binlog_max_cache_size.result
@@ -192,7 +192,7 @@ SET GLOBAL max_binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_cache_size= ORIGINAL_VALUE;
 SET GLOBAL max_binlog_stmt_cache_size= ORIGINAL_VALUE;
 SET GLOBAL binlog_stmt_cache_size= ORIGINAL_VALUE;
-include/stop_slave.inc
+include/stop_slave_io.inc
 include/start_slave.inc
 connection master;
 connection slave;

diff --git a/mysql-test/suite/rpl/r/rpl_stm_start_stop_slave.result b/mysql-test/suite/rpl/r/rpl_stm_start_stop_slave.result
@@ -109,7 +109,7 @@ START SLAVE;
 include/wait_for_slave_param.inc [Last_IO_Errno]
 Last_IO_Errno = '1236'
 Last_IO_Error = 'Got fatal error 1236 from master when reading data from binary log: 'Client requested master to start replication from impossible position; the first event 'master-bin.000001' at XXX, the last event read from 'master-bin.000001' at XXX, the last byte read from 'master-bin.000001' at XXX.''
-include/stop_slave.inc
+include/stop_slave_sql.inc
 RESET SLAVE;
 connection master;
 RESET MASTER;

diff --git a/mysql-test/suite/rpl/r/semisync_future-7591.result b/mysql-test/suite/rpl/r/semisync_future-7591.result
@@ -13,7 +13,7 @@ connection master;
 insert into t1 values (1);
 reset master;
 connection slave;
-include/stop_slave.inc
+include/stop_slave_sql.inc
 reset slave;
 include/start_slave.inc
 set global rpl_semi_sync_slave_enabled = OFF;

diff --git a/mysql-test/suite/rpl/t/mdev-31448_kill_ooo_finish_optimistic.test b/mysql-test/suite/rpl/t/mdev-31448_kill_ooo_finish_optimistic.test
@@ -54,7 +54,8 @@ drop table t2;
 --source include/start_slave.inc
 
 --echo # wait for T1
---let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(-1)' and  command LIKE 'Slave_worker';
+# Wildcard for `state` as it depends on whether WSREP is compiled in or not.
+--let $wait_condition= SELECT count(*)=1 FROM information_schema.processlist WHERE state LIKE 'Update_rows_log_event::find_row(%)' and  command LIKE 'Slave_worker';
 --source include/wait_condition.inc
 
 --echo # wait for T2
@@ -81,8 +82,9 @@ DROP TABLE t1;
 --connection slave
 --echo #
 --echo # Cleanup
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--let $rpl_allow_error= 1
+--source include/wait_for_slave_sql_to_stop.inc
+--source include/stop_slave_io.inc
 eval set @@global.slave_parallel_threads= $save_slave_parallel_threads;
 eval set @@global.slave_parallel_mode= $save_slave_parallel_mode;
 eval set @@global.innodb_lock_wait_timeout= $save_innodb_lock_wait_timeout;

diff --git a/mysql-test/suite/rpl/t/rpl_connection.test b/mysql-test/suite/rpl/t/rpl_connection.test
@@ -16,8 +16,7 @@ CHANGE MASTER TO MASTER_USER= '', MASTER_PASSWORD= '';
 START SLAVE;
 --let $slave_io_errno= 1045, 1593
 --source include/wait_for_slave_io_error.inc
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 
 CHANGE MASTER TO MASTER_USER= 'root', MASTER_PASSWORD= '';
 START SLAVE;

diff --git a/mysql-test/suite/rpl/t/rpl_domain_id_filter_io_crash.test b/mysql-test/suite/rpl/t/rpl_domain_id_filter_io_crash.test
@@ -149,8 +149,7 @@ connection slave;
 SELECT * FROM t1;
 
 SET @@global.debug_dbug=@saved_dbug;
---let $rpl_only_running_threads= 1
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 let $do_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Do_Domain_Ids, 1);
 let $ignore_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Ignore_Domain_Ids, 1);
 --echo DO_DOMAIN_IDS (BEFORE)     : $do_domain_ids_before
@@ -218,7 +217,7 @@ SELECT * FROM t1;
 
 SET @@global.debug_dbug=@saved_dbug;
 
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 let $do_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Do_Domain_Ids, 1);
 let $ignore_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Ignore_Domain_Ids, 1);
 --echo DO_DOMAIN_IDS (BEFORE)     : $do_domain_ids_before
@@ -286,7 +285,7 @@ SELECT * FROM t1;
 
 SET @@global.debug_dbug=@saved_dbug;
 
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 let $do_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Do_Domain_Ids, 1);
 let $ignore_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Ignore_Domain_Ids, 1);
 --echo DO_DOMAIN_IDS (BEFORE)     : $do_domain_ids_before
@@ -354,7 +353,7 @@ SELECT * FROM t1;
 
 SET @@global.debug_dbug=@saved_dbug;
 
---source include/stop_slave.inc
+--source include/stop_slave_sql.inc
 let $do_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Do_Domain_Ids, 1);
 let $ignore_domain_ids_before= query_get_value(SHOW SLAVE STATUS, Replicate_Ignore_Domain_Ids, 1);
 --echo DO_DOMAIN_IDS (BEFORE)     : $do_domain_ids_before