This is a milestone release, for use at your own risk. Significant development changes take place in milestone releases and you may encounter compatibility issues, such as data format changes that require attention in addition to the usual procedure of running mysql_upgrade. For example, you may find it necessary to dump your data with mysqldump before the upgrade and reload it afterward.
Audit Log Plugin Notes
Several changes were made to the audit log plugin for better compatibility with Oracle Audit Vault.
The format of the audit log file has changed:
elements written in the old format using attributes is
written in the new format using subelements.
The new format includes more information in
<AUDIT_RECORD> elements. Every
element includes a
providing a unique identifier. The
TIMESTAMP value includes time zone
information. Query records include
USER information, as well as
STATUS_CODE value differs from the
STATUS_CODE is 0 for success and 1 for
error, which is compatible with the EZ_collector consumer
for Audit Vault.
STATUS is the value of
mysql_errno() C API
function. This is 0 for success and nonzero for error, and
thus is not necessarily 1 for error.
Example of old
<AUDIT_RECORD TIMESTAMP="2013-04-15T15:27:27" NAME="Query" CONNECTION_ID="3" STATUS="0" SQLTEXT="SELECT 1" />
Example of new
<AUDIT_RECORD> <TIMESTAMP>2013-04-15T15:27:27 UTC</TIMESTAMP> <RECORD_ID>3998_2013-04-15T15:27:27</RECORD_ID> <NAME>Query</NAME> <CONNECTION_ID>3</CONNECTION_ID> <STATUS>0</STATUS> <STATUS_CODE>0</STATUS_CODE> <USER>root[root] @ localhost [127.0.0.1]</USER> <OS_LOGIN></OS_LOGIN> <HOST>localhost</HOST> <IP>127.0.0.1</IP> <COMMAND_CLASS>select</COMMAND_CLASS> <SQLTEXT>SELECT 1</SQLTEXT> </AUDIT_RECORD>
When the audit log plugin rotates the audit log file, it uses a
different file name format. For a log file named
audit.log, the plugin previously renamed
the file to
The plugin now renames the file to
to indicate that it is an XML file.
For information about the audit log plugin, see MySQL Enterprise Audit Log Plugin.
If you previously used an older version of the audit log plugin, use this procedure to avoid writing new-format log entries to an existing log file that contains old-format entries:
Stop the server.
Rename the current audit log file manually. This file will contain only old-format log entries.
Update the server and restart it. The audit log plugin will create a new log file, which will contain only new-format log entries.
The API for writing audit plugins has also changed. The
mysql_event_general structure has new members
to represent client host name and IP address, command class, and
external user. For more information, see
Writing Audit Plugins.
Functionality Added or Changed
Performance: String hashing overhead was reduced. This also improves performance for metadata locking, the table definition cache, and Performance Schema table I/O and file I/O instrumentation. (Bug #13944392)
PERFORMANCE_SCHEMA STATUS output used a mix of
attributes. These are now all
Similarly, the output used a mix of
size attributes. These are now all
Online index renaming is supported by
TABLE, which now includes a
INDEX clause, as shown in the following example:
ALTER TABLE t RENAME INDEX i1 TO
i1 is the current
name of the index and
i2 is the new name.
The result of “
ALTER TABLE t RENAME INDEX i1 TO
i2” would be a table with contents and
structure that is identical to the old version of
“t1” except for the index name, which is now
“i2” instead of “i1”.
ALTER TABLE t1 ALGORITHM=INPLACE, CHANGE COLUMN c1 c1 VARCHAR(255);
This is true as long as the number of length bytes required by a
VARCHAR column remains the same.
VARCHAR values of 0 to 255,
one length byte is required to encode the value. For
VARCHAR values of 256 bytes or
more, two length bytes are required. As a result, in-place
ALTER TABLE only supports
VARCHAR size from 0 to
255 bytes or increasing
size from a value equal to or greater than 256 bytes. In-place
ALTER TABLE does not support
VARCHAR size from less
than 256 bytes to a value equal to or greater than 256 bytes. In
this case, the number of required length bytes would change from
1 to 2, which is only supported by a table copy
InnoDB temporary table metadata is no longer stored to InnoDB
system tables. Instead, a new table,
users with a snapshot of active temporary tables. The table
contains metadata and reports on all user and system-created
temporary tables that are active within a given InnoDB instance.
The table is created when the first
SELECT statement is run against
Added support for spatial data types currently supported by
Prior to this release, InnoDB would store spatial data as binary
BLOB remains the underlying data type but
spatial data types are now mapped to a new InnoDB internal data
BLOB as the underlying data type,
a prefix index can still be used on the
GEOMETRY data column.
InnoDB: Added a separate tablespace for all non-compressed InnoDB temporary tables. The new tablespace is always recreated on server startup.
The new tablespace,
ibtmp1, is located in
the MySQL data directory (
datadir) by default. A newly
added configuration file option,
allows for a user-defined temporary data file path. For related
InnoDB Temporary Table Undo Logs.
HANDLER statements are now
supported with partitioned tables.
Auto_Position column has been added to the
output generated by
STATUS. The value of this column shows whether
replication autopositioning is in use. If autopositioning is
enabled—that is, if
1 was set by the last successful
CHANGE MASTER TO statement that
was executed on the slave—then the column's value is
1; if not, then the value is 0.
option for mysqlbinlog, which allows
mysqlbinlog to rewrite the names of databases
when playing back binary logs written using the row-based
logging format. Multiple rewrite rules can be created by
specifying the option multiple times.
GTID_SUBSET() were formerly
libmysqld only when it was built
with replication support. Now these functions are always
available when using this library, regardless of how it was
MySQL no longer uses the default OpenSSL compression. (Bug #16235681)
There is now a distinct error code
for the error sent by the server to a client authenticating with
an expired password.
In RPM packages built for Unbreakable Linux Network,
libmysqld.so now has a version number.
Error messages for
statement using a
ALGORITHM value not supported for the given
operation were very generic. The server now produces more
If a client with an expired password connected but
old_passwords was not the value
required to select the password hashing format appropriate for
the client account, there was no way for the client to determine
the proper value. Now the server automatically sets the session
appropriately for the account password.
mysqldump now supports an
--ignore-error option. The
option value is a comma-separated list of error numbers
specifying the errors to ignore during
mysqldump execution. If the
--force option is also given
to ignore all errors,
mysqlcheck has a new
--skip-database option. The
option value is the name of a database (case sensitive) for
which checks should be skipped.
mysql_upgrade uses this option to upgrade the
system tables in the mysql database before
tables in other databases: It upgrade the
mysql database, then all databases except the
mysql database. This avoids problems that can
occur if user tables are upgraded before the system tables.
variable was renamed to
Previously, on Linux the server failed to perform stack
backtrace attempts for versions of glibc
older than the current minimum supported version (2.3). Now on
such attempts the server displays a message that the
glibc version is too old to support
attached_condition information for
subqueries now includes
select# to indicate
the relative order of subquery execution.
The following changes were made to the sandbox mode that the server uses to handle client connections for accounts with expired passwords:
There is a new
system variable (default: enabled). This controls how the
server treats expired-password accounts.
Two flags were added to the C API client library:
mysql_real_connect(). Each flag enables a
client program to indicate whether it can handle sandbox
mode for accounts with expired passwords.
enabled for mysqltest unconditionally,
for mysql in interactive mode, and for
mysqladmin if the first command is
If a user attempted to access a nonexistent column for which the user had no access, the server returned an error indicating that the column did not exist. Now the server returns an error indicating that the user does not have privileges for the column, which provides no information about column existence. (Bug #19947, Bug #11745788)
system variable has been removed.
The mysql client now has a
--syslog option that causes
interactive statements to be sent to the system
syslog facility. Logging is suppressed for
statements that match the default “ignore” pattern
"*IDENTIFIED*:*PASSWORD*"), as well as
statements that match any patterns specified using the
--histignore option. For more
information, see mysql Logging.
ALTER TABLE now supports a
RENAME INDEX clause that renames an index.
The change is made in place without a table-copy operation.
The MySQL test suite mysql-test-run.sh
program now starts the server with
MyISAM as the default storage
engine. To maintain compatibility of test results with existing
result files, test cases were modified to add a line that
includes the the
file as necessary. In a future release, for those test cases not
MyISAM, that line will
be removed (so they run with
InnoDB) and test
results will be updated.
DROP TABLE statement for a
table using compression
could be slower than necessary, causing a stall for several
seconds. MySQL was unnecessarily decompressing
pages in the
buffer pool related to
the table as part of the
Incompatible Change; Partitioning:
Changes in the
KEY partitioning hashing
functions used with numeric, date and time,
SET columns in MySQL 5.5 makes
tables using partitioning or subpartitioning by
KEY on any of the affected column types and
created on a MySQL 5.5 or later server incompatible with a MySQL
5.1 server. This is because the partition IDs as calculated by a
MySQL 5.5 or later server almost certainly differ from those
calculated by a MySQL 5.1 server for the same table definition
and data as a result of the changes in these functions.
The principal changes in the
implementation in MySQL 5.5 resulting in this issue were as
follows: 1. The hash function used for numeric and date and time
columns changed from binary to character-based. 2. The base used
for hashing of
SET columns changed from
ci characters to binary.
The fix involves adding the capability in MySQL 5.5 and later to
choose which type of hashing to use for
partitioning, which is implemented with a new
ALGORITHM extension to the
BY KEY option for
PARTITION BY KEY ALGORITHM=1
([ causes the
server to use the hashing functions as implemented in MySQL 5.1;
ALGORITHM=2 causes the server to use
the hashing functions from MySQL 5.5 and later.
ALGORITHM=2 is the default. Using the
appropriate value for
ALGORITHM, you can
perform any of the following tasks:
KEY partitioned tables in MySQL
5.5 and later that are compatible with MySQL 5.1, using
CREATE TABLE ... PARTITION BY KEY ALGORITHM=1
KEY partitioned tables that
were created in MySQL 5.5 or later to become compatible with
MySQL 5.1, using
ALTER TABLE ... PARTITION BY KEY
KEY partitioned tables originally
created in MySQL 5.1 to use hashing as in MySQL 5.5 and
ALTER TABLE ... PARTITION BY KEY
Important: After such tables are
upgraded, they cannot be used any longer with MySQL 5.1
unless they are first downgraded again using
TABLE ... PARTITION BY KEY ALGORITHM=1 (...) on a
MySQL server supporting this option.
This syntax is not backward compatible, and causes errors in
older versions of the MySQL server. When generating
CREATE TABLE ...
PARTITION BY KEY statements,
mysqldump brackets any occurrence of
in conditional comments such that it is ignored by a MySQL
server whose version is not at least 5.5.31. An additional
consideration for upgrades is that MySQL 5.6 servers prior to
MySQL 5.6.11 do not ignore the
option in such statements when generated by a MySQL 5.5 server,
due to the that the conditional comments refer to version
5.5.31; in this case, you must edit the dump manually and remove
or comment out the option wherever it occurs before attempting
to load it into a MySQL 5.6.10 or earlier MySQL 5.6 server. This
is not an issue for dumps generated by MySQL 5.6.11 or later
version of mysqldump, where the version used
in such comments is 5.6.11. For more information, see
ALTER TABLE Partition Operations.
As part of this fix, a spurious assertion by
InnoDB that a deleted row had
previously been read, causing the server to assert on delete of
a row that the row was in the wrong partition, was also removed.
(Bug #14521864, Bug #66462, Bug #16093958, Bug #16274455)
References: See also Bug #11759782.
Important Change; Plugin API; Replication:
Because the behavior of the fulltext plugin may vary between
MySQL servers, it is not possible to guarantee that statements
using this plugin produce the same results on masters and
slaves. For this reason, statements depending on the fulltext
plugin are now marked as unsafe for statement-based logging.
This means that such statements are logged using row format when
binlog_format=MIXED, and cause
a warning to be generated when
(Bug #11756280, Bug #48183)
Important Change; Replication
This fix was reverted in MySQL 5.7.2. See the section called “Changes in MySQL 5.7.2 (2013-09-21, Milestone 12)”.
Executing a statement that performs an implicit commit but whose
changes are not logged when
gtid_next is set to any value
AUTOMATIC is not permitted. Now in
such cases, the statement fails with an error. This includes the
statements in the following list:
References: See also Bug #16484323.
Important Change; Replication:
The version number reported by mysqlbinlog
--version has been increased
(Bug #15894381, Bug #67643)
Important Change; Replication:
The lettercasing used for displaying UUIDs in global transaction
identifiers was inconsistent. Now, all GTID values use
lowercase, including those shown in the
Executed_Gtid_Set columns from the output of
SHOW SLAVE STATUS.
Important Note; Replication: Using row-based logging to replicate from a table to a same-named view led to a failure on the slave. Now, when using row-based logging, the target object type is checked prior to performing any DML, and an error is given if the target on the slave is not actually a table.
It remains possible to replicate from a table to a same-named view using statement-based logging.
(Bug #11752707, Bug #43975)
The setting for the
API node configuration parameter was ignored, and the default
value used instead.
MySQL Cluster: The recently added LCP fragment scan watchdog occasionally reported problems with LCP fragment scans having very high table id, fragment id, and row count values.
This was due to the watchdog not accounting for the time spent draining the backup buffer used to buffer rows before writing to the fragment checkpoint file.
Now, in the final stage of an LCP fragment scan, the watchdog switches from monitoring rows scanned to monitoring the buffer size in bytes. The buffer size should decrease as data is written to the file, after which the file should be promptly closed. (Bug #14680057)
MySQL Cluster: Job buffers act as the internal queues for work requests (signals) between block threads in ndbmtd and could be exhausted if too many signals are sent to a block thread.
Performing pushed joins in the
block can execute multiple branches of the query tree in
parallel, which means that the number of signals being sent can
increase as more branches are executed. If
DBSPJ execution cannot be completed before
the job buffers are filled, the data node can fail.
This problem could be identified by multiple instances of the message sleeploop 10!! in the cluster out log, possibly followed by job buffer full. If the job buffers overflowed more gradually, there could also be failures due to error 1205 (Lock wait timeout exceeded), shutdowns initiated by the watchdog timer, or other timeout related errors. These were due to the slowdown caused by the 'sleeploop'.
Normally up to a 1:4 fanout ratio between consumed and produced signals is permitted. However, since there can be a potentially unlimited number of rows returned from the scan (and multiple scans of this type executing in parallel), any ratio greater 1:1 in such cases makes it possible to overflow the job buffers.
The fix for this issue defers any lookup child which otherwise would have been executed in parallel with another is deferred, to resume when its parallel child completes one of its own requests. This restricts the fanout ratio for bushy scan-lookup joins to 1:1. (Bug #14709490)
References: See also Bug #14648712.
When parsing a delimited search string such as
“abc-def” in a full-text search,
InnoDB now uses the same word delimiters as
This fix improves code readability by addressing naming
inconsistencies for InnoDB
InnoDB: This fix disables a condition for extra splitting of clustered index leaf pages, on compressed tables. Extra page splitting was only done to reserve space for future updates, so that future page splits could be avoided. (Bug #16401801)
InnoDB tables, if a
KEY on a
(or prefix) was empty, index page compression could fail.
Status values in the
innodb_ft_config table would not
intended for internal configuration and should not be used for
statistical information purposes. To avoid confusion, column
values that are intended for internal use have been removed from
innodb_ft_config table. This
fix also removes the
innodb_ft_config table and other
internal full text search-related tables that were
(Bug #16409494, Bug #68502)
Improper testing of compatibility between the referencing and
ALTER TABLE ... ADD
FOREIGN key could cause a server exit.
InnoDB: Rollback did not include changes made to temporary tables by read-only transactions. (Bug #16310467)
InnoDB page-splitting algorithm could
For debug builds,
InnoDB status exporting was
subject to a race condition that could cause a server exit.
ALTER TABLE to set an
AUTO_INCREMENT column value to a
user-specified value, InnoDB would set the
AUTO_INCREMENT value to the user-specified
value even when the
AUTO_INCREMENT value is
greater than the user-specified value. This fix ensures that the
AUTO_INCREMENT value is set to the maximum of
the user-specified value and MAX(auto_increment_column)+1, which
is the expected behaviour.
InnoDB: Importing a tablespace with the configuration file present would not import the data file. This problem would occur when all pages are not flushed from the buffer pool after a table is altered using the copy and rename approach. This fix ensures that all pages are flushed from the buffer pool when a table is altered using the copy and rename approach. (Bug #16318052)
lock_validate function, which is only
present in debug builds, acquired and released mutexes to avoid
hogging them. This behavior introduced a window wherein changes
to the hash table could occur while code traversed the same set
of data. This fix updates
to collect all records for which locks must be validated,
releases mutexes, and runs a loop to validate record locks.
ALTER TABLE functions would
perform a check to see if InnoDB is in read-only mode
srv_read_only_mode=true). If InnoDB was in
read-only mode, the check would return a successful status and
do nothing else. This fix replaces
srv_read_only_mode check conditions with
InnoDB now aborts execution on Windows by calling the
abort() function directly, as it does on
When the InnoDB buffer pool is almost filled with 4KB compressed
pages, inserting into 16KB compact tables would cause 8KB
pages_free to increase, which could
potentially slow or stall inserts.
An assertion failure would occur in
MEM_BLOCK_MAGIC_N due to a race condition that
row_merge_read_clustered_index() returned an
InnoDB: This fix removes an unnecessary debug assertion related to page_hash locks which only affects debug builds. The debug assertion is no longer valid and should have been removed when hash_lock array was introduced in MySQL 5.6. (Bug #16263167)
Without warning, InnoDB would silently set
innodb-buffer-pool-instances to 1 if the
buffer pool size is less than 1GB. For example, if
innodb-buffer-pool-size is set to 200M and
innodb-buffer-pool-instances is set to 4,
InnoDB would silently set
innodb-buffer-pool-instances to 1. This fix
implements a warning message and new logic for
(Bug #16249500, Bug #61239)
InnoDB: When the primary key of a table includes a column prefix, and a full-text index is defined on the table, a full-text search resulted in an unnecessary warning being written to the error log. This fix suppresses the unnecessary warning. (Bug #16169411)
InnoDB: When InnoDB locking code was revised, a call to register lock waits was inadvertently removed. This fix adds the call back to the InnoDB locking code. (Bug #16208201)
A direct call to the
would cause a debug assertion.
In the case of LOCK WAIT for an insert in a foreign key table,
InnoDB could report a false
dictionary-changed error and cause the insert to fail rather
than being retried.
InnoDB: In some cases, deadlock detection did not work, resulting in sessions hanging waiting for a lock-wait timeout. (Bug #16169638)
ALTER TABLE on an
InnoDB table could fail to delete the
statistics for the old primary key from the
This fix updates InnoDB code in
handler0alter.cc to use
TABLE::key_info instead of both
Arithmetic underflow during page compression for
CREATE TABLE on an
InnoDB table could cause a server exit.
LOCK_TIME would not be logged correctly in
the slow query log.
LOCK_TIME did not account
for InnoDB row lock wait time.
For debug builds, online
TABLE operations for
could cause a server exit during table rebuilding.
In some cases, the
InnoDB purge coordinator
did not use all available purge threads, resulting in suboptimal
ALTER TABLE for
InnoDB tables was not fully atomic.
This fix replaces most uses of
UT_SORT_FUNCTION_BODY, an InnoDB recursive
merge sort, with the
from the C++ Standard Template Library (STL). The
std::sort() function requires less memory and
is faster due to in-line execution.
This fix implements a 256-byte boundary for extending a
VARCHAR column instead of 256-character
boundary. This change allows for in-place extension of a
VARCHAR column through an update of the data
This fix addresses unnecessary buffer pool lookups that would
occur while freeing blob pages, and implements a debug status
Creating numerous tables, each with a full-text search index,
could result in excessive memory consumption. This bug fix adds
a new configuration parameter,
which defines a global memory limit for full-text search
indexes. If the global limit is reached by an index operation, a
force sync is triggered.
(Bug #14834698, Bug #16817453)
InnoDB: This fix modifies InnoDB code to ensure that unused thread handles are closed when the thread exits, instead of leaving thread handles open until shutdown of mysqld on Windows. (Bug #14762796)
This fix removes unnecessary overhead by removing table locking
and disabling read view creation and MVCC when InnoDB is started
in read-only mode (
InnoDB: A regression introduced by the fix for Bug#14100254 would result in a “!BPAGE->FILE_PAGE_WAS_FREED” assertion. (Bug #14676249)
An error at the filesystem level, such as too many open files,
could cause an unhandled error during an
ALTER TABLE operation. The error
could be accompanied by Valgrind warnings, and by this assertion
Assertion `! is_set()' failed. mysqld got signal 6 ;
(Bug #14628410, Bug #16000909)
INNODB_SYNC_ARRAY_SIZE variable was
incorrectly allowed to be configured at runtime. As documented,
INNODB_SYNC_ARRAY_SIZE must be configured
when the MySQL instance is starting up, and cannot be changed
afterward. This fix changes
INNODB_SYNC_ARRAY_SIZE to a non-dynamic
variable, as intended.
The server could exit during an attempt by
InnoDB to reorganize or compress a compressed
secondary index page.
InnoDB: Full-text search (FTS) index savepoint information would not be set resulting in a severe error when attempting to rollback to the savepoint. (Bug #14639605, Bug #17456092)
A DML operation performed while a
TABLE operation waits for pending I/O operations on
the tablespace to complete would result in a deadlock.
InnoDB: Attempting to unninstall the InnoDB memcached Plugin while the plugin is still installing caused the Mysql server to terminate. While the plugin deamon thread was still initializing, plugin variables were not yet set and the uninstall process could not cleanup resources. This fix adds a variable to indicate initialization status. If initialization is incomplete, the uninstall process will wait. (Bug #14279541)
If the value of
innodb_force_recovery was less
than 6, opening a corrupted table might loop forever if a
corrupted page was read when calculating statistics for the
table. Information about the corrupted page was written
repeatedly to the error log, possibly causing a disk space
issue. The fix causes the server to halt after a fixed number of
failed attempts to read the page. To troubleshoot such a
corruption issue, set
(Bug #14147491, Bug #65469)
When printing out long semaphore wait diagnostics,
sync_array_cell_print() ran into a
segmentation violation (SEGV) caused by a race condition. This
fix addresses the race condition by allowing the cell to be
freed while it is being printed.
InnoDB: Killing a query caused an InnoDB assertion failure when the same table (cursor) instance was used again. This is the result of a regression error introduced by the fix for Bug#14704286. The fix introduced a check to handle kill signals for long running queries but the cursor was not restored to the proper state. (Bug #68051, Bug #16088883)
InnoDB: On startup, InnoDB reported a message on 64-bit Linux and 64-bit Windows systems stating that the CPU does not support crc32 instructions. On Windows, InnoDB does not use crc32 instructions even if supported by the CPU. This fix revises the wording of the message and implements a check for availability of crc32 instructions. (Bug #68035, Bug #16075806)
InnoDB: The length of internally generated foreign key names was not checked. If internally generated foreign key names were over the 64 character limit, this resulted in invalid DDL from SHOW CREATE TABLE. This fix checks the length of internally generated foreign key names and reports an error message if the limit is exceeded. (Bug #44541, Bug #11753153)
This fix removes left-over prototype code for
srv_parse_log_group_home_dirs, and related
(Bug #68133, Bug #16198764)
Attempting to replace the default InnoDB full-text search (FTS)
stopword list by creating an InnoDB table with the same
would result in an error.
TABLE revealed that the new InnoDB table was created
CHARSET=utf8. The InnoDB FTS stopword
table validity check only supported latin1. This fix extends the
validity check for all supported character sets.
(Bug #68450, Bug #16373868)
A query on a table partitioned by range and using
TO_DAYS() as a partitioing
function always included the first partition of the table when
pruning. This happened regardless of the range employed in the
BETWEEN clause of such a query.
(Bug #15843818, Bug #49754)
Partition pruning is now enabled for tables using a storage
engine that provides automatic partitioning, such as the
NDB storage engine, but which are
explicitly partitioned. Previously, pruning was disabled for all
tables using such a storage engine, whether or not the tables
had explicitly defined partitions.
In addition, as part of this fix, explicit partition selection
is now disabled for tables using a storage engine (such as
NDB) that provides automatic partitioning.
References: See also Bug #14672885.
TABLE ... DROP PARTITION against a view caused the
server to crash, rather than fail with an error as expected.
A query result was not sorted if both
ORDER BY were
used and the underlying table was partitioned.
Inserting any number of rows into an
ARCHIVE table that used more than
1000 partitions and then attempting to drop the table caused the
MySQL Server to fail.
(Bug #13819630, Bug #64580)
Replication; Linux; Microsoft Windows:
Replication failed between a Linux master using
lower_case_table_names set to 0
and a Windows slave having
lower_case_table_names set to 2, after a
replicated table was opened on the slave; in addition,
was required afterwards to see which updates had actually been
applied on the slave. This was because
lower_case_table_names was checked only to
see whether it was equal to 1 prior to forcing a conversion of
replicated database object names to lower case for checking the
table cache. Now in such cases,
checked to see whether it is set to a nonzero value.
Replication: When using GTIDs and binary log auto-positioning, the master had to scan all binary logs whenever the slave reconnected (due to reasons such as I/O thread failure or a change of master) before it could send any events to slave. Now, the master starts from the oldest binary log that contains any GTID not found on the slave. (Bug #16340322, Bug #68386)
Replication: When the server version of the master was greater than or equal to 10, replication to a slave having a lower server version failed. (Bug #16237051, Bug #68187)
Replication: When replicating to a MySQL 5.6 master to an older slave, Error 1193 (ER_UNKNOWN_SYSTEM_VARIABLE) was logged with a message such as Unknown system variable 'SERVER_UUID' on master, maybe it is a *VERY OLD MASTER*. This message has been improved to include more information, similar to this one: Unknown system variable 'SERVER_UUID' on master. A probable cause is that the variable is not supported on the master (version: 5.5.31), even though it is on the slave (version: 5.6.11). (Bug #16216404, Bug #68164)
A zero-length name for a user variable (such as
@``) was incorrectly considered to be a sign
of data or network corruption when reading from the binary log.
(Bug #16200555, Bug #68135)
SHOW RELAYLOG EVENTS at a slave where
no relay log file is present returned the following incorrect
error message: "
Error when executing command SHOW
BINLOG EVENTS: Could not find target log." The error
message text has been changed to: "
Error when executing
command SHOW RELAYLOG EVENTS: Could not find target
When MTS is on and transactions are being applied, the slave
coordinator would hang when encountering a checksum error on a
transaction event. This was due to a deadlock situation in which
the coordinator assumed a normal stop while a worker waited for
the coordinator to dispatch more events. For debug builds, the
problem appeared as an assertion failure, which was due to the
coordinator not setting
encountering an error.
The print format specifier for the
was incorrectly defined as a signed 32-bit integer with a range
of -2144783647 to 2144783648. This fix changes the
server_id integer type to an unsigned 32-bit
integer type, with a range of 0 to 4294967295, which is the
documented range for the
mysqlbinlog can connect to a remote server
and read its binary logs. In MySQL 5.6 and later, this tool can
also wait for the server to generate and send additional events,
in practice behaving like a slave connecting to a master. In
cases where the server sent a heartbeat,
mysqlbinlog was unable to handle it properly.
As a consequence, mysqlbinlog failed at this
point, without reading any more events from the server. To fix
this problem, mysqlbinlog now ignores any
binary log events of type
that it receives.
STOP SLAVE could cause a deadlock
when issued concurrently with a statement such as
SHOW STATUS that retrieved the
values for one or more of the status variables
(Bug #16088188, Bug #67545)
References: See also Bug #16088114.
--replicate-* options (see
Replication Slave Options and Variables) could in some cases
lead to a memory leak on the slave.
(Bug #16056813, Bug #67983)
`) characters were not always
handled correctly in internally generated SQL statements, which
could sometimes lead to errors on the slave.
(Bug #16084594, Bug #68045)
References: This bug is a regression of Bug #14548159, Bug #66550.
The session-level value for
gtid_next was incorrectly reset
on the slave for all rollbacks, which meant that GTIDs could be
lost for multi-statement transactions, causing the slave to stop
with an ER_GTID_NEXT_TYPE_UNDEFINED_GROUP
error. Now this is done only when a complete transaction is
being rolled back, or when
autocommit is enabled.
In order to provision or to restore a server using GTIDs, it is
possible to set
a given GTID set listing the transactions that were imported.
This operation requires that the global
gtid_purged server system
variables are empty. (This is done in order to avoid the
possibility of overriding server-generated GTIDs.)
The error message GTID_PURGED can only be set when GTID_EXECUTED is empty that was raised when this requirement was not met could be confusing or misleading because it did not specify the scope of the affected variables. To prevent this from happening, error messages that refer to variables relating to GTIDs now specify the scope of any such variables when they do so. (Bug #16084426, Bug #68038)
In certain cases, the dump thread could send a heartbeat out of
synchronisation with format description events. One of the
effects of this issue what that, after provisioning a new server
from a backup data directory and setting
--gtid-mode=ON and enabling
CHANGE MASTER TO Syntax),
replication failed to start, with the error Read
invalid event from master.... The same problem could
also cause GTID-based replication to fail due to skipped events
following a unplanned shutdown of the master.
Replication: In some cases, when the slave could not recognize the server version of the master, this could cause the slave to fail. (Bug #16056365)
Replication: Dropping a table that includes non-regular ASCII characters in the table name caused a replication failure. The parser converted the table name into standard charset characters and stored the table name in the table_name variable. When the drop table query was regenerated using the table_name variable, the table name was not converted back to the original charset.
Additionally, table and database names with 64 characters caused an assert failure. The assert required the table or database name to be less than 128 characters. Latin characters require two-bytes each, which requires an assert condition of less than or equal to 128 bits.
The fix includes a new function to convert tables names back to the original charset, and a correction to the assert condition allowing table and database names be less than or equal to 128 bits. (Bug #16066637)
Replication failed when a replicated
DATA statement inserted rows into a view.
(Bug #15993712, Bug #67878)
Replication: When using GTID-based replication, and whenever a transaction was executed on the master but was not sent to the slave because the slave already had a transaction with that ID, semisynchrononous replication timed out. One case in which this could happen was during a failover operation where the new master started behind the new slave. (Bug #15985893)
An unnecessary flush to disk performed after every transaction
FILE as the replication info
repository type could degrade performance. Now this is done only
when both data and relay log info is stored in (transactional)
Replication: When a slave read a table map event from the binary log, it assumed that the metadata size was always less than twice the column count of the table in use, which failed when the event contained the wrong value for this field. (Bug #15830022)
Replication: When reading row log events from the binary log, the slave assumed that these events were always valid; because of this, an event having an invalid binary log offset could cause the slave to crash. Now in such cases, the slave fails gracefully, and an error is reported, if any of the fields in a given row event are invalid. (Bug #15829568)
Table IDs used in replication were defined as type
ulong on the master and
uint on the slave. In addition, the maximum
value for table IDs in binary log events is 6 bytes
(281474976710655). This combination of factors led to the
Data could be lost on the slave when a table was assigned an
ID greater than
Table IDs greater than 281474976710655 were written to the binary log as 281474976710655.
This led to a stopped slave when the slave encountered two tables having the same table ID.
To fix these problems, IDs are now defined by both master and
slave as type
ulonglong but constrained to a
range of 0 to 281474976710655, restarting from 0 when it exceeds
(Bug #14801955, Bug #67352)
MASTER_POS_WAIT() could hang or
return -1 due to invalid updates by the slave SQL thread when
transactions were skipped by the GTID protocol.
References: See also Bug #15927032.
Replication: Trying to execute a Stop event on a multi-threaded slave could cause unwanted updates to the relay log, leading the slave to lose synchronization with the master. (Bug #14737388)
Replication: Internal objects used for relay log information were only partially deleted before freeing their memory. (Bug #14677824)
Replication: When the server starts, it checks whether semisynchronous replication has been enabled without a lock, and, if so, it takes the lock, then tests again. Disabling semisynchronous replication following the first of the these tests, but prior to the second one, could lead to a crash of the server. (Bug #14511533, Bug #66411)
Replication: It was possible in certain cases—immediately after detecting an EOF in the dump thread read event loop, and before deciding whether to change to a new binary log file—for new events to be written to the binary log before this decision was made. If log rotation occurred at this time, any events that occurred following EOF detection were dropped, resulting in loss of data. Now in such cases, steps are taken to make sure that all events are processed before allowing the log rotation to take place. (Bug #13545447, Bug #67929)
References: See also Bug #16016886.
It was possible for the
MASTER_POS_WAIT() function to
return prematurely following a
MASTER TO statement that updated the
RELAY_LOG_NAME. This could happen because
CHANGE MASTER TO did not update the master
log position in such cases, causing
MASTER_POS_WAIT() to read an invalid log
position and to return immediately.
To fix this problem, the master log position is flagged as
invalid until the position is set to a valid value when the SQL
thread reads the first event, after which it is flagged as
valid. Functions such as
now defer any comparison with the master log position until a
valid value can be obtained (that is, after the first event
CHANGE MASTER TO statement has
(Bug #11766010, Bug #59037)
Replication: If the disk becomes full while writing to the binary log, the server hangs until space is freed up manually. It was possible after this was done for the MySQL server to fail, due to an internal status value being set when not needed. Now in such cases, rather than trying to set this status, a warning is written in the error log instead. (Bug #11753923, Bug #45449)
Replication: The binary log and relay log files used the name of the PID file instead of the host name as the basename. (Bug #11753843, Bug #45359)
Microsoft Windows: In Shared Memory mode, the MySQL Server could crash when receiving requests from multiple threads. (Bug #13934876)
RPM packages were missing the
Windows MSI installers for MySQL 5.7 had a 5.6 upgrade code, not a 5.7 upgrade code. (Bug #16445344)
SHOW ENGINE PERFORMANCE_SCHEMA STATUS could
report incorrect memory-allocation values when the correct
values exceeded 4GB.
Performance Schema statement tokenization overhead was reduced. (Bug #16382260)
A long database name in a
statement could cause the server to exit.
The server could exit if a prepared statement attempted to create a table using the name of an existing view while an SQL handler was opened. (Bug #16385711)
For debug builds, checking of password constraints could raise an assertion for statements that updated passwords. (Bug #16289303)
BUILD-CMAKE file in MySQL distributions
was updated with the correct URL for CMake
The optimizer's attempt to remove redundant subquery clauses
raised an assertion when executing a prepared statement with a
subquery in the
ON clause of a join in a
References: This bug is a regression of Bug #15875919.
values could cause an assertion to be raised.
Some aggregate queries attempted to allocate excessive memory. (Bug #16343992)
Incorrect results were returned if a query contained a subquery
IN clause which contained an
XOR operation in the
For debug builds, an assertion could be raised if a statement
failed with autocommit enabled just before an
XA START statement
Conversion of numeric values to
BIT could yield unexpected
HAVING clauses were rejected as
Fixed warnings when compiling with XCode 4.6. Fixed warnings
when compiling when the
isoctal macro was already defined in the
(Bug #16265300, Bug #60911, Bug #12407384)
Queries using range predicates that were evaluated using the LooseScan semi-join strategy could return duplicate rows. (Bug #16221623)
References: This bug is a regression of Bug #14728469.
For upgrade operations, RPM packages produced unnecessary errors
about being unable to access
In the range optimizer, an index merge failure could cause a server exit. (Bug #16241773)
For debug builds, the server could exit due to incorrect
calculation of applicable indexes for a join that involved
The Performance Schema could return incorrect values for the
PROCESSLIST_INFO column of the
Invocation of the range optimizer for a
select caused the server to exit.
mysql_config --libs displayed incorrect output. (Bug #16200717)
CREATE TABLE (...
statement for which
col_name TIMESTAMP DEFAULT
CURRENT_TIMESTAMP ...) ... SELECT
SELECT did not provide a value for the
TIMESTAMP column, that column was set to
'0000-00-00 00:00:00', not the current timestamp.
yaSSL did not perform proper padding checks, but instead examined only the last byte of plaintext and used it to determine how many bytes to remove. (Bug #16218104)
GROUP BY WITH ROLLUP in a prepared
statement could cause the server to exit.
References to the unused
SIGNAL_WITH_VIO_SHUTDOWN macro in the
CMake files were removed.
system variable to an inappropriate value could cause the server
Directory name manipulation could result in stack overflow on Mac OS X and Windows. (Bug #16066243)
For debug builds, if the server was started with binary logging
EVENTS from within a stored procedure raised an
The query parser leaked memory for some syntax errors. (Bug #16040022)
With the thread pool plugin enabled, large numbers of connections could lead to a Valgrind panic or failure of clients to be able to connect. (Bug #16088658, Bug #16196591)
The server executed
FORMAT=JSON for some malformed queries improperly.
Performance Schema instrumentation was missing for slave worker threads. (Bug #16083949)
If the error for a failed
INDEX statement index within a stored program was
processed by a condition handler, a malformed packet and
“Command out of sync” error occurred.
test database contained a
dummy.bak file that prevented
DATABASE from working. This file is no longer
included. Also, a
db.opt file is now included
that contains these lines:
Setting a system variable to
cause the server to exit.
SET PASSWORD and
GRANT ... IDENTIFIED
BY have no effect on the password of a user who is
authenticated using an authentication plugin that accesses
passwords stored externally to the mysql.user table. But
attempts to change the password of such a user produced no
warning, leading to the impression that the password had been
changed when it was not. Now MySQL issues an
warning to indicate that the attempt was ignored.
For debug builds, creating an InnoDB table in strict SQL mode that violated the maximum key length limit caused the server to exit. (Bug #16035659)
PREPARE statement using certain
combinations of stored functions and user variables caused the
server to exit.
WITH_CSV_STORAGE_ENGINE in the server source code were
removed because the
MERGE storage engine plugins are mandatory.
mysys library string-formatting routine
could mishandle width specifiers.
could set connection character set system variables to values
ucs2 that are not permitted.
During shutdown, the server could attempt to lock an uninitialized mutex. (Bug #16016493)
permitted invalid plugin values, and did not always set the
old_passwords system variable to a value
appropriate for the named plugin.
For debug builds, executing a statement within a trigger or stored function that caused an implicit commit raised an assertion. (Bug #15985318)
Under some circumstances, mysql --secure-auth permitted passwords to be sent to the server using the old (pre-4.1) hashing format. (Bug #15977433)
With index condition pushdown enabled, queries for which the pushed-down condition contained no columns in the used index could be slow. (Bug #15896009)
Table creation operations added entries to the Performance
file_instances table, but
these were not always removed for table drop operations.
In special cases, the optimizer did not consider indexes that
were applicable to query processing, resulting in potentially
suboptimal execution and incorrect
(Bug #15849135, Bug #16094171)
A query with an
ORDER BY clause ordering by an outer
column of type
BLOB that is not in the select
list caused an assertion to fire.
References: See also Bug #14728142.
Queries in the query cache involving a given table were
incorrectly invalidated if a
of the same name was dropped.
The optimizer could return nonmatching records for queries that
ref access on string
References: See also Bug #14682735.
CREATE SERVER due to a
missing or read-only
resulted in a memory leak.
Enabling the query cache during high client contention could cause the server to exit. (Bug #14727815)
Enabling the slow query log at runtime when access permissions on the log file prevented the server from writing to it caused the server to exit. (Bug #14711347)
Table removal could fail and cause the server to exit for very long file names. (Bug #14581920)
The server sometimes failed to respect
MAX_CONNECTIONS_PER_HOUR limits on user
The server could access the
facility while closing temporary tables during connection
shutdown, after the facility had been cleaned up, leading to an
assertion being raised.
The optimizer could return incorrect results after transforming
IN subquery with aggregate functions to an
If the optimizer calculated a row count of zero for the inner table of an outer join, it could not determine proper ordering for the following tables. (Bug #14628746)
When a client program loses the connection to the MySQL server
or if the server begins a shutdown after the client has executed
mysql_stmt_prepare(), the next
mysql_stmt_prepare() returns an
error (as expected) but subsequent
mysql_stmt_execute() calls crash
Previously, if multiple
were given, mysql_config_editor ignored all
but the last one. Now multiple
options result in an error.
If MySQL server was started with options to enable the general query log or slow query log, but access permissions on the log file prevented the server from writing to it, the server started with an error message indicating that logging was off and that the server needed to be restarted after the problem was corrected. This was incorrect because it is also possible to set the logging variables again at runtime (without a restart) after correcting the problem. The error message now indicates this possibility. (Bug #14512467)
For debug builds, creating a
inside a trigger caused the server to exit.
system variables indicate key files for the
sha256_password authentication plugin, but
the server failed to properly check whether the key files were
valid. Now in the event that either key file is invalid, the
server logs an error and exits.
cause the server to exit. This syntax is now prohibited because
SET context there is no column name and
the statement returns
COM_CHANGE_USER command in the
client/server protocol did not properly use the character set
number in the command packet, leading to incorrect character set
conversion of other values in the packet.
with a locale and a very large number could cause the server to
For debug builds, improper handling for
AUTO_INCREMENT value overflow could cause the
server to exit.
Certain plugin-related conditions can make a user account unusable:
The account requires an authentication plugin that is not loaded.
The account requires the
authentication plugin but the server was started with
neither SSL nor RSA enabled as required by this plugin.
The server now checks those conditions by default and produces
warnings for unusable accounts. This checking slows down server
FLUSH PRIVILEGES, so it is
made optional by means of the new
validate_user_plugins system variable. This
variable is enabled by default, but if you do not require the
additional checking, you can disable it at startup to avoid the
(Bug #13010061, Bug #14506305)
Passing an unknown time zone specification to
CONVERT_TZ() resulted in a memory
glibc-2.2.5.patch files in the
Docs directory of MySQL distributions have
mysql_install_db did not escape
'_' in the host name for statements written
to the grant tables.
The server could exit if built to permit a maximum number of indexes per table larger than 64.
In the course of fixing this problem, a
CMake option was added to permit building the
server to support a larger maximum number of indexes per table.
The default is 64. The maximum is 255. Values smaller than 64
are ignored and the default of 64 is used.
Source code in the
mysys library for the
my_free_lock memory-locking APIs was never
used and has been removed.
(Bug #54662, Bug #11762107)
mysqld_safe used the nonportable
-e test construct.
(Bug #67976, Bug #16046140)
An out-of-memory condition could occur while handling an out-of-memory error, leading to recursion in error handling. (Bug #49514, Bug #11757464)
The optimizer used loose index scan for some queries for which this access method is inapplicable. (Bug #42785, Bug #11751794)
If a dump file contained a view with one character set and collation defined on a view with a different character set and collation, attempts to restore the dump file failed with an “illegal mix of collations” error. (Bug #65382, Bug #14117025)
If the server failed to read
could exit with a segmentation fault.
(Bug #53393, Bug #11760944)
REPLACE() function produced
incorrect results when a user variable was supplied as an
argument and the operation was performed on multiple rows.
(Bug #49271, Bug #11757250)
The output for
SHOW CREATE VIEW
could vary depending on the
(Bug #34553, Bug #11747931)
UNION type conversion could
incorrectly turn unsigned values into signed values.
(Bug #49003, Bug #11757005)
If one thread was rebuilding a result for the query cache, other threads in the middle of using the previous result could fail to discard the old result properly. For debug builds, this raised an assertion. (Bug #66781, Bug #14631798)
View access in low memory conditions could raise a debugging assertion. (Bug #39307, Bug #11749556)
Attempts to create a trigger for which a trigger with the same
action time and event already existed resulted in an
rather than an
(Bug #67357, Bug #14801721)
Queries with many values in a
were slow due to inclusion of debugging code in non-debugging
(Bug #68046, Bug #16078212)
References: See also Bug #58731, Bug #11765737.
When a view definition contained a special character in the
SEPARATOR clause of the
function, mysqldump created an invalid view
definition that produced an error when the dump file was
(Bug #60920, Bug #12395512)
max_connections to a
value less than the current number of open connections caused
the server to exit.
(Bug #44100, Bug #11752803)
Some table I/O performed by the server when calling a storage engine were missing from the statistics collected by the Performance Schema. (Bug #68180, Bug #16222630)
For debug builds, some queries with
SELECT ... FROM
DUAL nested subqueries raised an assertion.
(Bug #60305, Bug #11827369)
Nonspatial indexes only support exact-match lookups for spatial
columns, but the optimizer incorrectly used
range access in some cases,
leading to incorrect results.
(Bug #67889, Bug #15993693)
If mysql is built with the bundled
libedit library, the library is built as
static code, to avoid linking to a different dynamic version at
runtime. Dynamic linking could result in use of a different,
incompatible version and a segmentation fault.
(Bug #68231, Bug #16296509)
SLEEP() produced no warning or
NULL or negative arguments. Now it
produces a warning, or an error in strict SQL mode.
(Bug #67548, Bug #15859462)
command options now are exposed at runtime as the
system variables. Their values can be examined using
SHOW VARIABLES. The variables are
dynamic, so their values can can be set at runtime. (The options
were actually replaced by the system
variables, but as system variables can be set at server startup,
no option functionality is lost.)
(Bug #59860, Bug #11766693)
For arguments with fractional seconds greater than six decimals,
SEC_TO_TIME() truncated, rather
than rounding as it should have.
(Bug #68061, Bug #16093024)
MySQL failed to build if configured with
(Bug #67018, Bug #16342793)
If a table had rows in the
table, dropping the table did not remove those rows.
(Bug #67283, Bug #14779330)
resulted in link errors due to selection of the incorrect
(Bug #68277, Bug #16284051)
ALTER TABLE inserted
col_name TIMESTAMP DEFAULT
CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP
0000-00-00 00:00:00 rather than the current
timestamp if the alteration was done in place rather than by
making a table copy.
(Bug #68040, Bug #16076089)
If the server was started with
ALTER USER ... PASSWORD EXPIRE caused the
server to exit.
(Bug #68300, Bug #16295905)
CMake did not check whether the system
zlib had certain functions required for
MySQL, resulting in build errors. Now it checks and falls back
to the bundled
zlib if the functions are
(Bug #65856, Bug #14300733)
mysql_install_db did not work in Solaris 10 sparse root zones. (Bug #68117, Bug #16197860)
EXPLAIN DELETE and
listed all indexes, not just the applicable indexes.
(Bug #67830, Bug #15972078)
The Perl version of mysql_install_db mishandled some error messages. (Bug #68118, Bug #16197542)
ORDER BY and
LIMIT could lead to incorrect results for
(Bug #68458, Bug #16383173)
ORDER BY clause following an
IN subquery could cause duplicate rows to be
(Bug #68330, Bug #16308085)
If XA support was activated by multiple storage engines, the server would exit. (Bug #47134, Bug #11755370)
A statement with an aggregated, nongrouped outer query and an
aggregated, nongrouped subquery in the
list could return incorrect results.
(Bug #68372, Bug #16325175)
SQL_NO_CACHE keyword is supposed to
prevent the server from checking the query cache to see whether
the query result is already cached, and to prevent it from
caching the query result. However, the query cache check was
suppressed only if
SQL_NO_CACHE was preceded
and followed by space characters. (For example, the server
checked the cache if the keyword was followed by a newline.) Now
the parser requires that the preceding and following characters
be whitespace characters, not spaces.
(Bug #64164, Bug #13641256)
NULL into a
TIMESTAMP NOT NULL column now produces an
error (as it already did for other
data types), instead of inserting the current timestamp.
(Bug #68472, Bug #16394472)
KILL to kill a statement
in another session could in some cases cause that session to
return an incorrect error code.
(Bug #45679, Bug #11754124)
If a column is declared as
NOT NULL, it is
not permitted to insert
NULL into the column
or update it to
NULL. However, this
constraint was enforced even if there was a
BEFORE UPDATE trigger)
that set the column to a non-
NULL value. Now
the constraint is checked at the end of the statement, per the
(Bug #6295, Bug #11744964)
On Windows, the
system variable did not accurately reflect the error log file
name in some cases. For example, if the server was started
--log-error, the default is to
the data directory, but
log_error should be nonblank and reflect
the log file name in all cases, on all platforms. The value is
stderr if the server does not write error
messages to a file and sends them to the console (standard error
output) instead. In particular, on Windows,
--console overrides use of an error log and
sends error messages to the console, so
log_error will be set to
(Bug #8307, Bug #11745049)