This patch adds functionality to row-based replication to ensure the
slave's column sizes are >= to that of the master.
It also includes some refactoring for the code from WL#3228.
Fixing tests and results to work when replicating to fewer columns on
slave than on master. One test that previously should fail, now works,
and some log positions have changed as a result of adding metadata to
the events.
This patch adds the ability to store extra field metadata in the table
map event. This data can include pack_length() or field_lenght() for
fields such as CHAR or VARCHAR enabling developers to add code that
can check for compatibilty between master and slave columns. More
importantly, the extra field metadata can be used to store data from the
master correctly should a VARCHAR field on the master be <= 255 bytes
while the same field on the slave is > 255 bytes.
The patch also includes the needed changes to unpack to ensure that data
which is smaller on the master can be unpacked correctly on the slave.
WL#3915 : (NDB) master's cols > slave
Slave starts accepting and handling rows of master's tables which have more columns.
The most important part of implementation is how to caclulate the amount of bytes to
skip for unknown by slave column.
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.
minor fixes to appease pushbuild.
---
Merge sin.intern.azundris.com:/home/tnurnberg/22540/50-22540
into sin.intern.azundris.com:/home/tnurnberg/22540/51-22540
previous correction didn't. make sure "tail" is fixed up
when filling cache several times; rework formulae.
---
Merge sin.intern.azundris.com:/home/tnurnberg/22540/50-22540
into sin.intern.azundris.com:/home/tnurnberg/22540/51-22540
fix binlog-writing so that end_log_pos is given correctly even
within transactions for both SHOW BINLOG and SHOW MASTER STATUS,
that is as absolute values (from log start) rather than relative
values (from transaction's start).
---
Merge sin.intern.azundris.com:/home/tnurnberg/22540/50-22540
into sin.intern.azundris.com:/home/tnurnberg/22540/51-22540
---
Merge tnurnberg@bk-internal.mysql.com:/home/bk/mysql-5.1-maint
into sin.intern.azundris.com:/home/tnurnberg/22540/51-22540
New test cases for generic apply status table
rpl_ndb_dd_advance.test:
Updated test case to not run sooo long
show_binlog_using_logname.inc, tpcb_disk_data.inc:
New support file
rpl_ndb_dd_advance.result:
Updated result
tpcb.inc:
New support file for Updated test case to not run sooo long
Adding new fields Last_{IO,SQL}_Errno and Last_{IO,SQL}_Error to output
of SHOW SLAVE STATUS to hold errors from I/O and SQL thread respectively.
Old fields Last_Error and Last_Errno are aliases for Last_SQL_Error and
Last_SQL_Errno respectively.
Fields are added last to output of SHOW SLAVE STATUS to allow old applications
to use the same positional arguments into the row, while allowing new
application to benefit from the added information.
In addition, some new error codes are added (especially for the I/O
thread) to be able to provide sensible error message.
The patch for WL 1563 added a new duplicate key error message so that the
key name could be provided instead of the key number. But the error code
for the new message was used even though that did not need to change.
This could cause unnecessary problems for applications that used the old
ER_DUP_ENTRY error code to detect duplicate key errors.
This patch corrects a bug involving a LOAD DATA INFILE operation on a
transactional table. It corrects a problem in the error handler by moving
the transactional table check and autocommit_or_rollback operation to the
end of the error handler.
The problem was an assert was thrown after the operation completed. The
assert found a non-sunk event in the transaction cache. The events in the
transaction cache were added after commit_or_rollack and thereafter nothing
removed them.
An additional test case was added to detect this
condition.
---
Merge pippilotta.erinye.com:/shared/home/df/mysql/build/mysql-5.0-build-work-vanilla-building
into pippilotta.erinye.com:/shared/home/df/mysql/build/mysql-5.1-build-work-vanilla-building
---
Fix test cases to pass for a plain ./configure && make build. This includes disabling two test cases when certain features are not present in the server. We're not losing coverage from this because these features are usually present, and disabling them here only serves the purpose to make the test cases work in the unlikely case that they aren't.
---
fixes
(Removes some warnings about UNIX_TIMESTAMP from the slave.err logs)
Marked federated_server as a '--big-test'
Change error in net_clear to 'Note', as it interfered with mysql-test-run.
Adding an event that can be used to denote that an incident occured
on the master. The event can be used to denote a gap in the replication
stream, but can also be used to denote other incidents.
In addition, the injector interface is extended with functions to
generate an incident event. The function will also rotate the binary
log after generating an incident event to get a fresh binary log.
There was hanging at binlog_commit by a thread executing autocommit query.
The hang appeared to be due to an overly condtion for early return
from binlog_commit introduced by bug#20265 fix.
Fixed with reverting the logic back to 5.0 version.
field does not work
Fix to prevent MyISAM from reading data from NULL BLOB.
Fix to make record comparison independent of values of unused bits in
record.
Updating binlog positions in tests.
1. Fixes within the testscripts (affects rpl_ddl.test and rpl_ndb_ddl.test)
- slave connection is only an observer (-> AUTOCOMMIT = 0)
This removes the problem with the hanging test around DROP DATABASE (NDB). The hanging test around DROP DATABASE is a difference to InnoDB/MyISAm behaviour but fare
away of a clear bug. IMHO this behaviour does not violate the SQL standard and should be therefore simply accepted.
- removal of wrong comments
- CREATE/DROP TEMPORARY TABLE must not cause implicit commit of the current transaction.
NDB behaves here correct and InnoDB/Falcon wrong.
- Add a missing connection slave
- Reenable the test rpl_ndb_ddl.
2. Disable rpl_ddl.test because of Bug#26418.
3. Reenable rpl_ndb_ddl.test
4. Improvements (affect rpl_ddl.test and rpl_ndb_ddl.test)
- Better + extended comments which should prevent that somebody accidently destroys the logics of the test
- Replace SELECT's printing comments by "--echo" (decreases the number of auxiliary SQL commands)
- Remove the need for include/rpl_stmt_seq2.inc (was mostly redundant to rpl_stmt_seq.inc)
- Remove extra/rpl_tests/rpl_ndb_ddl.test (corrected extra/rpl_tests/rpl_ddl.test is sufficient)
- Shift assignment of values to $show_binlog, $manipulate (variables useful for debugging) into the toplevel scripts
- The temporary tables get now their storage engine from the variable $temp_engine_type. (more deterministic testing conditions)
- Add additional protocol line if the connection is switched (was partially missing)
- Add two DML commands for comparison purposes
With this patch, statements that change metadata (in the mysql database)
is logged as statements, while normal changes (e.g., using INSERT, DELETE,
and/or UPDATE) is logged according to the format in effect.
The log tables (i.e., general_log and slow_log) are not replicated at all.
With this patch, the following statements are replicated as statements:
GRANT, REVOKE (ALL), CREATE USER, DROP USER, and RENAME USER.
1)
BUG#25507 "multi-row insert delayed + auto increment causes
duplicate key entries on slave" (two concurrrent connections doing
multi-row INSERT DELAYED to insert into an auto_increment column,
caused replication slave to stop with "duplicate key error" (and
binlog was wrong), and BUG#26116 "If multi-row INSERT
DELAYED has errors, statement-based binlogging breaks" (the binlog
was not accounting for all rows inserted, or slave could stop).
The fix is that: in statement-based binlogging, a multi-row INSERT
DELAYED is silently converted to a non-delayed INSERT.
This is supposed to not affect many 5.1 users as in 5.1, the default
binlog format is "mixed", which does not have the bug (the bug is
only with binlog_format=STATEMENT).
We should document how the system delayed_insert thread decides of
its binlog format (which is not modified by this patch):
this decision is taken when the thread is created
and holds until it is terminated (is not affected by any later change
via SET GLOBAL BINLOG_FORMAT). It is also not affected by the binlog
format of the connection which issues INSERT DELAYED (this binlog
format does not affect how the row will be binlogged).
If one wants to change the binlog format of its server with SET
GLOBAL BINLOG_FORMAT, it should do FLUSH TABLES to be sure all
delayed_insert threads terminate and thus new threads are created,
taking into account the new format.
2)
BUG#24432
"INSERT... ON DUPLICATE KEY UPDATE skips auto_increment values".
When in an INSERT ON DUPLICATE KEY UPDATE, using
an autoincrement column, we inserted some autogenerated values and
also updated some rows, some autogenerated values were not used
(for example, even if 10 was the largest autoinc value in the table
at the start of the statement, 12 could be the first autogenerated
value inserted by the statement, instead of 11). One autogenerated
value was lost per updated row. Led to exhausting the range of the
autoincrement column faster.
Bug introduced by fix of BUG#20188; present since 5.0.24 and 5.1.12.
This bug breaks replication from a pre-5.0.24/pre-5.1.12 master.
But the present bugfix, as it makes INSERT ON DUP KEY UPDATE
behave like pre-5.0.24/pre-5.1.12, breaks replication from a
[5.0.24,5.0.34]/[5.1.12,5.1.15]
master to a fixed (5.0.36/5.1.16) slave! To warn users against this when
they upgrade their slave, as agreed with the support team, we add
code for a fixed slave to detect that it is connected to a buggy
master in a situation (INSERT ON DUP KEY UPDATE into autoinc column)
likely to break replication, in which case it cannot replicate so
stops and prints a message to the slave's error log and to SHOW SLAVE
STATUS.
For 5.0.36->[5.0.24,5.0.34] replication or 5.1.16->[5.1.12,5.1.15]
replication we cannot warn as master
does not know the slave's version (but we always recommended to users
to have slave at least as new as master).
As agreed with support, I have asked for an alert to be put into
the MySQL Network Monitoring and Advisory Service.
3) note that I'll re-enable rpl_insert_id as soon as 5.1-rpl gets
the changes from the main 5.1.
does not work): Changing packed row format to only include null bits
for those columns that are present in the row as well as writing BIT
columns in a storage engine-independent format.
The change in row format is incompatible with the previous format and a
slave will not be able to read the new events.
The test uses show binlog event which is not deterministic due to the single insert delayed
query can generate up to number of inserted rows row-events pair (table_map + Write_row)
The solution is to leave the current binlogging behaviour as it is and change
the test as spliting arguments of insert delayed query. Note, that such fix was applied
earlier for binlog_insert_delayed.test :
https://intranet.mysql.com/secure/apps/irclog.php?channel=22&start_time=2006-09-27
There are no tests with insert delayed and show binlog events combination requiring
this fix.
Field_bit::set_default() did not check the bit_len, hence used the undefined
bit_ptr, causing a crash. The patch adds a check that bit_len > 0 before
following the bit_ptr.
Bug#23831 deadlock not noticed
RBR bug in that when replicated msta (multi-statement-trans-action) deadlocks
with a local at write row event or gets timed-out, the event handler did not return
the correct error code.
Wrong error code stops slave sql thread instead of to proceed with
rollback and replay.
The correct code is typed in error log and stored for error handling rotine
to conduct rollback and replay of the transaction. The handling for the rbr
remains the same as for the sbr events.
Particularly, timed-out transaction still is rolled back - look at the related bugs.
Bug #22027 CREATE TABLE IF NOT EXISTS SELECT logged improperly with row-based binlog
post-merge fix, also making results free from server's version number.
and #22762: create talbe if not exists like a-temp-table binlogged w/o "if not exists"
These are rbr bugs.
store_create_info, which generates the create statement e.g for binlogging,
did not account a lex option HA_LEX_CREATE_IF_NOT_EXISTS.
The fix makes the generated query to include the parent's statement option
if
that was supplied.
ChangeSet@1.2309.1.12, 2006-09-12 15:42:13+02:00, guilhem@gbichot3.local +14 -0
Fixing problems I identified in my auto_increment work pushed in July
(as part of the auto_increment cleanup of WL #3146; ...
The problem is in that show binlog events in indeterministic, row events can be compressed,
so that 2 seconds original delay does not guard from inconsistency.
We syncronize test's current inserted rows counter with system insert delayed thread
per each query.
From another side there is no requirement for binlog to be event per row and then
to verify if binlog has recorded what was recently inserted is better
via reading from it instead of 'show binlog events'.
(as part of the auto_increment cleanup of WL#3146; let's not be
sad, that monster push still removed serious bugs):
one problem with INSERT DELAYED (unexpected interval releases),
one with stored functions (wrong auto_inc binlogging).
These bugs were not released.
The following is an excerption from the WL.
1. Change so that MIXED is default format
1.1 to change the default for command line --binlog-format
1.2 to alter global_system_variables.binlog_format calculation
basing on command line --binlog-format parameter and
its default.
2. Change test suite so that more testing is done by MIXED format.
2.1 to check if there are test cases requiring --binlog-foramt=statement via
`source include/have_binlog_format_statement.inc' and affected by
altering the latter to be "mixed".
2.2 to check the content of such vulnerable cases to find if
extending to the mixed does not modify results. In that case simply
substitute source arguments as explained.
2.3 if a test in mixed mode deals with features triggering
row-binlogging then if necessary we can switch explicitly
to statement mode or create another test to run with
non-recommended STATEMENT mode
Particullarily, extracting INSERT DELAYED
binlogging subtest for statement mode is performed, and
the snippet is moved into a separate test file.
Note that since now all three modes verify this use case
through 3 different tests.
No changes in item 3 of HLD appeared to be needed.
We now reset the THD members related to auto_increment+binlog in
MYSQL_LOG::write(). This is better than in THD::cleanup_after_query(),
which was not able to distinguish between SELECT myfunc1(),myfunc2()
and INSERT INTO t SELECT myfunc1(),myfunc2() from a binlogging point
of view.
Rows_log_event::exec_event() now calls lex_start() instead of
mysql_init_query() because the latter now does too much (it resets
the binlog format).
this is a cleanup patch for our current auto_increment handling:
new names for auto_increment variables in THD, new methods to manipulate them
(see sql_class.h), some move into handler::, causing less backup/restore
work when executing substatements.
This makes the logic hopefully clearer, less work is is needed in
mysql_insert().
By cleaning up, using different variables for different purposes (instead
of one for 3 things...), we fix those bugs, which someone may want to fix
in 5.0 too:
BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate
statement-based"
BUG#20341 "stored function inserting into one auto_increment puts bad
data in slave"
BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE"
(now if a row is updated, LAST_INSERT_ID() will return its id)
and re-fixes:
BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT"
(already fixed differently by Ramil in 4.1)
Test of documented behaviour of mysql_insert_id() (there was no test).
The behaviour changes introduced are:
- LAST_INSERT_ID() now returns "the first autogenerated auto_increment value
successfully inserted", instead of "the first autogenerated auto_increment
value if any row was successfully inserted", see auto_increment.test.
Same for mysql_insert_id(), see mysql_client_test.c.
- LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY
UPDATE, see auto_increment.test. Same for mysql_insert_id(), see
mysql_client_test.c.
- LAST_INSERT_ID() does not change if no autogenerated value was successfully
inserted (it used to then be 0), see auto_increment.test.
- if in INSERT SELECT no autogenerated value was successfully inserted,
mysql_insert_id() now returns the id of the last inserted row (it already
did this for INSERT VALUES), see mysql_client_test.c.
- if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X
(it already did this for INSERT VALUES), see mysql_client_test.c.
- NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE,
the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID
influences not only the first row now.
Additionally, when unlocking a table we check that the thread is not keeping
a next_insert_id (as the table is unlocked that id is potentially out-of-date);
forgetting about this next_insert_id is done in a new
handler::ha_release_auto_increment().
Finally we prepare for engines capable of reserving finite-length intervals
of auto_increment values: we store such intervals in THD. The next step
(to be done by the replication team in 5.1) is to read those intervals from
THD and actually store them in the statement-based binary log. NDB
will be a good engine to test that.
