mariadb/mysql-test/extra/rpl_tests/rpl_log.test
unknown b9c18aead4 Bug#19259 rpl_ndb_dd_partitions failed on Solaris
Actually, this testcase will fail generally on all testing platforms.
The bugs come from the inconsistent bitmap between rpl master and slave.

In log_event.cc, the n_bits of m_cols and m_cols_ai are intialized with octal-ceiling 
m_width, in fact, their n_bits should be equal to m_width.
Wrong n_bits will cause bitmap_bits_set() get incorrect value in unpack_row()
in rpl_record.cc, 
then an assertion in unpack_row() will fail and crash sql thread.
  DBUG_ASSERT(null_ptr == row_data + master_null_byte_count);

Meanwhile, because of binlog_prepare_pending_rows_event() changed with correct
m_cols, some results of specific testcases should be updated:
binlog_multi_engine.test
ndb_binlog_multi.test
rpl_ndb_dd_partitions.test
rpl_ndb_log.test
rpl_truncate_7ndb.test
rpl_truncate_7ndb_2.test

In addition, to ensure rows replication correct between master and slave after the patch, 
two 'select * from t1' are added in extra/rpl_tests/rpl_log.test, and some testcases include 
rpl_log.test, therefore, the results of these testcases should be updated likewise:
rpl_stm_log.test
rpl_row_log.test
rpl_ndb_log.test
rpl_row_log_innodb.test

Totally, results of nine testcases are updated.


include/my_bitmap.h:
  add declaration for create_last_word_mask(), since we need to use it in log_event.cc
mysql-test/extra/rpl_tests/rpl_log.test:
  add 'select * from t1' both on master and slave to ensure the replication consistent after patched
mysql-test/r/binlog_multi_engine.result:
  update result
  
  the following is commented by Andrei, 
  The differences can be explained.
  Look at
  lines of THD::binlog_prepare_pending_rows_event
  
    if (!pending)
         ...
        !bitmap_cmp(pending->get_cols(), cols))
  
  The row event disappears because after correcting master side Write_rows_log_event's m_col bitmap initialization, 
  this event has the same m_cols as the preceding one and thereafter they got glued into one bigger size event. 
  This is the correct behaviour now.
mysql-test/r/ndb_binlog_multi.result:
  update result
  same reason with binlog_multi_engine for Write_rows_log_event
mysql-test/r/rpl_ndb_dd_partitions.result:
  update result,
  reasons:
  1. tablespace description format changed
  2. hex(b1) from 1 to 0 because there is a update sentence in include/rpl_multi_engine3.inc
  UPDATE t1 SET b1 = 0, t="2006-02-22" WHERE id = 412;
mysql-test/r/rpl_ndb_log.result:
  update result
  Write_rows event disapper for same reason with binlog_multi_engine ;
  In addition, because add new select in extra/rpl_testsrpl_log.test, 
  corresponding new results are accompanying.
mysql-test/r/rpl_row_log.result:
  because add new select in extra/rpl_testsrpl_log.test, 
  corresponding new results are accompanying.
mysql-test/r/rpl_row_log_innodb.result:
  because add new select in extra/rpl_testsrpl_log.test, 
  corresponding new results are accompanying.
mysql-test/r/rpl_stm_log.result:
  because add new select in extra/rpl_testsrpl_log.test, 
  corresponding new results are accompanying.
mysql-test/r/rpl_truncate_7ndb.result:
  update result
  same reason with binlog_multi_engine for Write_rows_log_event
mysql-test/r/rpl_truncate_7ndb_2.result:
  update result
  same reason with binlog_multi_engine for Write_rows_log_event
  
  And for the change
  -master-bin.000001  4   Format_desc  1   102 Server ver: SERVER_VERSION, Binlog ver: 4
  +master-bin.000001  4   Format_desc  1   106 Server ver: SERVER_VERSION, Binlog ver
  
  It's  okay as FormatDescription event matured for 4 bytes since the last time the results had been recorded.
mysql-test/t/disabled.def:
  resume this test case
sql/log_event.cc:
  initialize m_cols' n_bits with m_width instead of octal-round(m_width);
  initialize m_cols_ai' n_bits with m_width instead of octal-round(m_width);
  after memcpy(), call create_last_word_mask() to clear extra bits in bitmap to ensure safety
mysql-test/t/rpl_ndb_dd_partitions-master.opt:
  add --new=true passed to mysqld
mysql-test/t/rpl_ndb_dd_partitions-slave.opt:
  add --new=true passed to mysqld
2007-07-12 15:19:29 +08:00

152 lines
5.8 KiB
Text

# Requires statement logging
# Clean up old slave's binlogs.
# The slave is started with --log-slave-updates
# and this test does SHOW BINLOG EVENTS on the slave's
# binlog. But previous tests can influence the current test's
# binlog (e.g. a temporary table in the previous test has not
# been explicitly deleted, and at the beginning of the current
# test the slave immediately writes DROP TEMPORARY TABLE this_old_table).
# We wait for the slave to have written all he wants to the binlog
# (otherwise RESET MASTER may come too early).
save_master_pos;
connection slave;
sync_with_master;
stop slave;
reset master;
reset slave;
# We are going to read the slave's binlog which contains file_id (for some LOAD
# DATA INFILE); to make it repeatable (not influenced by other tests), we need
# to stop and start the slave, to be sure file_id will start from 1.
# This can be done with 'server_stop slave', but
# this would require the manager, so most of the time the test will be skipped
# :(
# To workaround this, I (Guilhem) add a (empty) rpl_log-slave.opt (because when
# mysql-test-run finds such a file it restarts the slave before doing the
# test). That's not very elegant but I could find no better way, sorry.
let $VERSION=`select version()`;
connection master;
reset master;
eval create table t1(n int not null auto_increment primary key)ENGINE=$engine_type;
insert into t1 values (NULL);
drop table t1;
eval create table t1 (word char(20) not null)ENGINE=$engine_type;
load data infile '../std_data_ln/words.dat' into table t1 ignore 1 lines;
select count(*) from t1;
--replace_result $VERSION VERSION
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events;
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events from 106 limit 1;
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events from 106 limit 2;
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events from 106 limit 2,1;
flush logs;
# We need an extra update before doing save_master_pos.
# Otherwise, an unlikely scenario may occur:
# * When the master's binlog_dump thread reads the end of master-bin.001,
# it send the rotate event which is at this end, plus a fake rotate event
# because it's starting to read a new binlog.
# save_master_pos will record the position of the first of the two rotate
# (because the fake one is not in the master's binlog anyway).
# * Later the slave waits for the position of the first rotate event,
# and it may quickly stop (in 'slave stop') without having received the fake
# one.
# So, depending on a few milliseconds, we end up with 2 rotate events in the
# relay log or one, which influences the output of SHOW SLAVE STATUS, making
# it not predictable and causing random test failures.
# To make it predictable, we do a useless update now, but which has the
# interest of making the slave catch both rotate events.
eval create table t3 (a int)ENGINE=$engine_type;
# Sync slave and force it to start on another binary log
#show rows of t1 on master
connection master;
select * from t1 order by 1 asc;
save_master_pos;
connection slave;
# Note that the above 'slave start' will cause a 3rd rotate event (a fake one)
# to go into the relay log (the master always sends a fake one when replication
# starts).
start slave;
let $result_pattern= '%127.0.0.1%root%master-bin.000002%slave-relay-bin.000005%Yes%Yes%0%0%None%';
--source include/wait_slave_status.inc
sync_with_master;
#check t1 on slave to ensure whether it's identical with on master
select * from t1 order by 1 asc;
flush logs;
stop slave;
connection master;
# Create some entries for second log
eval create table t2 (n int)ENGINE=$engine_type;
insert into t2 values (1);
source include/show_binlog_events.inc;
--replace_result $VERSION VERSION
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events in 'master-bin.000002';
show binary logs;
save_master_pos;
connection slave;
start slave;
sync_with_master;
show binary logs;
--replace_result $MASTER_MYPORT MASTER_PORT $VERSION VERSION
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events in 'slave-bin.000001' from 4;
--replace_result $MASTER_MYPORT MASTER_PORT $VERSION VERSION
--replace_column 2 # 5 #
--replace_regex /\/\* xid=.* \*\//\/* XID *\// /table_id: [0-9]+/table_id: #/
show binlog events in 'slave-bin.000002' from 4;
source include/show_slave_status.inc;
# Need to recode the following
#show new master for slave with master_log_file='master-bin.000001' and master_log_pos=4 and master_server_id=1;
#show new master for slave with master_log_file='master-bin.000001' and master_log_pos=79 and master_server_id=1;
#show new master for slave with master_log_file='master-bin.000001' and master_log_pos=311 and master_server_id=1;
#show new master for slave with master_log_file='master-bin.000002' and master_log_pos=4 and master_server_id=1;
#show new master for slave with master_log_file='master-bin.000002' and master_log_pos=122 and master_server_id=1;
--error 1220
show binlog events in 'slave-bin.000005' from 4;
connection master;
# The table drops caused Cluster Replication wrapper to fail as event ID would never be the same.# Moving drops here.
DROP TABLE t1;
DROP TABLE t2;
DROP TABLE t3;
#
# Bug #6880: LAST_INSERT_ID() within a statement
#
create table t1(a int auto_increment primary key, b int);
insert into t1 values (NULL, 1);
reset master;
set insert_id=5;
insert into t1 values (NULL, last_insert_id()), (NULL, last_insert_id());
source include/show_binlog_events.inc;
select * from t1;
drop table t1;
# End of 4.1 tests
sync_with_master;