diff options
| -rw-r--r-- | mysql-test/suite/rpl/r/rpl_semi_sync_slave_reply_fail.result | 40 | ||||
| -rw-r--r-- | mysql-test/suite/rpl/t/rpl_semi_sync_slave_reply_fail.test | 87 | ||||
| -rw-r--r-- | sql/slave.cc | 13 |
3 files changed, 134 insertions, 6 deletions
diff --git a/mysql-test/suite/rpl/r/rpl_semi_sync_slave_reply_fail.result b/mysql-test/suite/rpl/r/rpl_semi_sync_slave_reply_fail.result new file mode 100644 index 00000000000..6b39b296cdf --- /dev/null +++ b/mysql-test/suite/rpl/r/rpl_semi_sync_slave_reply_fail.result @@ -0,0 +1,40 @@ +include/master-slave.inc +[connection master] +connection slave; +include/stop_slave.inc +connection master; +call mtr.add_suppression("Timeout waiting for reply of binlog*"); +set global rpl_semi_sync_master_enabled = ON; +SET @@GLOBAL.rpl_semi_sync_master_timeout=100; +create table t1 (i int); +connection slave; +set global rpl_semi_sync_slave_enabled = ON; +CALL mtr.add_suppression("Semi-sync slave net_flush*"); +SET @save_debug= @@global.debug; +SET GLOBAL debug_dbug="+d,semislave_failed_net_flush"; +include/start_slave.inc +connection master; +connection slave; +"Assert that the net_fulsh() reply failed is present in slave error log. +FOUND 1 /Semi-sync slave net_flush\(\) reply failed/ in mysqld.2.err +"Assert that Slave IO thread is up and running." +SHOW STATUS LIKE 'Slave_running'; +Variable_name Value +Slave_running ON +Slave_IO_Running= Yes +"Clear the network failure simulation." +SET GLOBAL debug_dbug= @save_debug; +connection master; +insert into t1 values (10); +connection slave; +connection slave; +# Compare the tables on master and slave. +include/diff_tables.inc [master:t1, slave:t1] +connection master; +drop table t1; +connection slave; +set global rpl_semi_sync_slave_enabled = OFF; +connection master; +set global rpl_semi_sync_master_enabled = OFF; +SET @@GLOBAL.rpl_semi_sync_master_timeout = 10000; +include/rpl_end.inc diff --git a/mysql-test/suite/rpl/t/rpl_semi_sync_slave_reply_fail.test b/mysql-test/suite/rpl/t/rpl_semi_sync_slave_reply_fail.test new file mode 100644 index 00000000000..f0eb474f00e --- /dev/null +++ b/mysql-test/suite/rpl/t/rpl_semi_sync_slave_reply_fail.test @@ -0,0 +1,87 @@ +# ==== Purpose ==== +# +# Test verifies that slave IO thread doesn't report an error, when slave fails +# to send an acknowledgment to master with semi sync replication in use. +# +# ==== Implementation ==== +# +# Steps: +# 0 - Have semi synchronous replication in use. +# 1 - Enable a debug simulation point which simulates network flush failure +# at the time of slave reply operation. +# 2 - Do some operation on master and wait for it to be replicated. Master +# will timeout waiting for reply from slave. +# 3 - Check the slave error log for appropriate error message regarding +# net_flush operation failure. +# 4 - Remove the debug simulation and do some more DML operations on master +# and wait for them to be replicated. +# 5 - Slave will be able to replicate and data is consistent on both master +# and slave. Semi sync will be automatically turned on. +# +# ==== References ==== +# +# MDEV-20217: Semi_sync: Last_IO_Error: Fatal error: Failed to run +# 'after_queue_event' hook +# +--source include/have_debug.inc +--source include/master-slave.inc + +--connection slave +--source include/stop_slave.inc + +--connection master +call mtr.add_suppression("Timeout waiting for reply of binlog*"); +--let $sav_timeout_master=`SELECT @@GLOBAL.rpl_semi_sync_master_timeout` +set global rpl_semi_sync_master_enabled = ON; +SET @@GLOBAL.rpl_semi_sync_master_timeout=100; +create table t1 (i int); + +--connection slave +set global rpl_semi_sync_slave_enabled = ON; +CALL mtr.add_suppression("Semi-sync slave net_flush*"); +SET @save_debug= @@global.debug; +SET GLOBAL debug_dbug="+d,semislave_failed_net_flush"; +--source include/start_slave.inc + +--connection master +--sync_slave_with_master + +# Check error log for correct messages. +let $log_error_= `SELECT @@GLOBAL.log_error`; +if(!$log_error_) +{ + # MySQL Server on windows is started with --console and thus + # does not know the location of its .err log, use default location + let $log_error_ = $MYSQLTEST_VARDIR/log/mysqld.2.err; +} +--echo "Assert that the net_fulsh() reply failed is present in slave error log. +--let SEARCH_FILE=$log_error_ +--let SEARCH_PATTERN=Semi-sync slave net_flush\(\) reply failed +--source include/search_pattern_in_file.inc + +--echo "Assert that Slave IO thread is up and running." +SHOW STATUS LIKE 'Slave_running'; +let $status= query_get_value("show slave status", Slave_IO_Running, 1); +echo Slave_IO_Running= $status; + +--echo "Clear the network failure simulation." +SET GLOBAL debug_dbug= @save_debug; + +--connection master +insert into t1 values (10); +--sync_slave_with_master + +--connection slave +--echo # Compare the tables on master and slave. +--let $diff_tables= master:t1, slave:t1 +--source include/diff_tables.inc + +--connection master +drop table t1; +--sync_slave_with_master +set global rpl_semi_sync_slave_enabled = OFF; + +--connection master +set global rpl_semi_sync_master_enabled = OFF; +--eval SET @@GLOBAL.rpl_semi_sync_master_timeout = $sav_timeout_master +--source include/rpl_end.inc diff --git a/sql/slave.cc b/sql/slave.cc index 4007f323a6e..165aa20b1ee 100644 --- a/sql/slave.cc +++ b/sql/slave.cc @@ -4911,13 +4911,14 @@ Stopping slave I/O thread due to out-of-memory error from master"); goto err; } - if (rpl_semi_sync_slave_status && (mi->semi_ack & SEMI_SYNC_NEED_ACK) && - repl_semisync_slave.slave_reply(mi)) + if (rpl_semi_sync_slave_status && (mi->semi_ack & SEMI_SYNC_NEED_ACK)) { - mi->report(ERROR_LEVEL, ER_SLAVE_FATAL_ERROR, NULL, - ER_THD(thd, ER_SLAVE_FATAL_ERROR), - "Failed to run 'after_queue_event' hook"); - goto err; + /* + We deliberately ignore the error in slave_reply, such error should + not cause the slave IO thread to stop, and the error messages are + already reported. + */ + (void)repl_semisync_slave.slave_reply(mi); } if (mi->using_gtid == Master_info::USE_GTID_NO && |