Web lists-archives.com

MySQL Community Server 5.7.26 has been released





Dear MySQL users,

MySQL Server 5.7.26, a new version of the popular Open Source Database
Management System, has been released. MySQL 5.7.26 is recommended for
use on production systems.

For an overview of what's new in MySQL 5.7, please see

http://dev.mysql.com/doc/refman/5.7/en/mysql-nutshell.html

For information on installing MySQL 5.7.26 on new servers, please see
the MySQL installation documentation at

http://dev.mysql.com/doc/refman/5.7/en/installing.html

MySQL Server 5.7.26 is available in source and binary form for a number
of platforms from our download pages at

http://dev.mysql.com/downloads/mysql/

MySQL Server 5.7.26 is also available from our repository for Linux
platforms, go here for details:

http://dev.mysql.com/downloads/repo/ <http://dev.mysql.com/downloads/repo/>

Windows packages are available via the Installer for Windows or .ZIP
(no-install) packages for more advanced needs. The point and click
configuration wizards and all MySQL products are available in the
unified Installer for Windows:

http://dev.mysql.com/downloads/installer/

5.7.26 also comes with a web installer as an alternative to the full
installer.

The web installer doesn't come bundled with any actual products and
instead relies on download-on-demand to fetch only the products you
choose to install. This makes the initial download much smaller but
increases install time as the individual products will need to be
downloaded.

We welcome and appreciate your feedback, bug reports, bug fixes,
patches, etc.:

http://bugs.mysql.com/report.php <http://bugs.mysql.com/report.php>

The following link lists the changes in the MySQL 5.7 since the release
of MySQL 5.7.25. It may also be viewed online at

http://dev.mysql.com/doc/relnotes/mysql/5.7/en/news-5-7-26.html

Enjoy!


==============================================================================

Changes in MySQL 5.7.26 (2019-04-25, General Availability)

   Beginning with MySQL 5.7.26, Oracle no longer provides binaries for
   SUSE 11.

     * Security Notes

     * Bugs Fixed

Security Notes


     * The linked OpenSSL library for the MySQL Commercial
       Server has been updated to version 1.0.2r. Issues fixed in the
       new OpenSSL version are described at
http://www.openssl.org/news/vulnerabilities.html. This change
       does not affect the Oracle-produced MySQL Community build of
       MySQL Server, which uses the yaSSL library instead. (Bug
       #28988091)

Bugs Fixed


     * Important Note: The libevent library included with the
       MySQL Server was upgraded to version 2.1.8. (Bug #28207237, Bug
       #29041505, Bug #29055011)

     * InnoDB: Optimized InnoDB internal temporary tables did
       not support in-place UPDATE operations, which caused the number
       of delete-marked records to increase continuously.  The large
       number of delete-marked records could cause longer than expected
       query execution times. (Bug #29207450)

     * InnoDB: The base column information for a generated
       column was not stored. (Bug #29021730)

     * InnoDB: Assertion code related to the innodb_flush_method
       O_DIRECT_NO_FSYNC setting was no longer valid due to a recent
       modification to that setting. Assertion code was revised. (Bug
       #29007731) References: See also: Bug #27309336.

     * InnoDB: Memory leaks discovered in the innochecksum
       utility were removed. (Bug #28917614, Bug #93164)

     * InnoDB: A DDL operation that followed a failed attempt to
       create an index on a virtual column resulted in an assertion
       failure. (Bug #28825718)

     * InnoDB: A Linux AIO handler function failed to check if
       completed I/O events succeeded. Thanks to Wei Zhao for the
       contribution. (Bug #27850600, Bug #90402)

     * InnoDB: A function called by a CREATE TABLE thread
       attempted to access a table object after it was freed by a
       background thread.  Thanks to Yan Huang for the patch. (Bug
       #27373959, Bug #89126)

     * InnoDB: Two sessions concurrently executing an INSERT ...
       ON DUPLICATE KEY UPDATE operation generated a deadlock.  During
       partial rollback of a tuple, another session could update it. The
       fix for this bug reverts fixes for Bug #11758237, Bug #17604730,
       and Bug #20040791. (Bug #25966845)

     * InnoDB: When the method used to access a joined table was
       const, InnoDB attempted to unlock the matching row multiple
       times. (Bug #20939184)

     * InnoDB: The INDEX_LENGTH value in
       INFORMATION_SCHEMA.TABLES was not updated when adding an index.
       (Bug #19811005)

     * Partitioning: An AUTO_INCREMENT key added to a
       partitioned table by an ALTER TABLE statement using
       ALGORITHM=INPLACE restarted on each partition. (Bug #92241, Bug
       #28573894)

     * Replication: If the WAIT_FOR_EXECUTED_GTID_SET() function
       was used with a timeout value including a fractional part (for
       example, 1.5), an error in the casting logic meant that the
       timeout was rounded down to the nearest whole second, and to zero
       for values less than 1 second (for example, 0.1). The casting
       logic has now been corrected so that the timeout value is applied
       as originally specified with no rounding. Thanks to Dirkjan
       Bussink for the contribution. (Bug #29324564, Bug #94247)

     * Replication: Depending on the value of
       group_replication_exit_state_action, the behavior of members
       exiting a group was not consistent. To harmonize the behavior of
       members exiting the group regardless of the error scenario, now
       when a member with group_replication_exit_state_action=READ_ONLY
       exits the group unintentionally, the super_read_only mode that
       the member had when started is restored. This makes the behavior
       consistent with that of a member with
       group_replication_exit_state_action=ABORT_SERVER. (Bug #28971639,
       Bug #28526591)

     * Replication: Whenever you use
       group_replication_allow_local_disjoint_gtids_join, which is
       deprecated, a log message is added. (Bug #28971624)

     * Replication: On overloaded servers there was a
       possibility that when a member joined the group, the
       VIEW_CHANGE_LOG_EVENT event which marks that point was not logged
       in the correct place. This could lead to errors in the data
       transfer to the newly joining server and data divergence. Now,
       the VIEW_CHANGE_LOG_EVENT event is logged in the correct place in
       the binary log. In addition, warnings are logged about the delay
       in logging the event. (Bug #28971594)

     * Replication: If an applier thread was stopped while it
       was in the process of opening a table, no error was set, which
       could result in a segmentation fault or assertion depending on
       the build type. Error handling is now correctly activated in this
       situation. (Bug #28864557)

     * Replication: With GTIDs in use on the server, the master
       info log on a replication slave was being synchronized every time
       the master skipped a transaction using the auto-skip function.
       The process ends with a dummy heartbeat which is sent to the
       slave and caused a forced flush to the log, and this could have a
       large cumulative impact on the write load on the slave. The same
       issue could occur in a circular replication topology with events
       that originated from the same server and were therefore ignored,
       which were also handled by the slave with a forced flush to the
       log. The slave handling code has now been changed to remove the
       forced flush for heartbeat events and for ignored events received
       through circular replication, so that the master info log is only
       synchronized when appropriate (for example, when a CHANGE MASTER
       statement is issued, or the binary log is rotated). (Bug
       #28815555, Bug #85158)

     * Replication: In a replication group configured in
       single-primary mode (group_replication_single_primary_mode=ON,
       which is the default), if severe network delays affected the
       group, it was possible for the primary and the secondaries to
       reach different decisions on a transaction, which could lead to
       divergence in the gtid_executed sets on the members. The issue
       has now been fixed. (Bug #28768550, Bug #28966455, Bug #92690)

     * Replication: GRANT statements that were written to the
       binary log were logged incorrectly in some cases, which could
       result in a GRANT statement that executed successfully on the
       master causing an error on the replication slave. (Bug #28643405,
       Bug #29155451, Bug #93750)

     * Replication: If a storage engine has the capability to
       log in STATEMENT format but not in ROW format, when binlog_format
       is set to STATEMENT, an unsafe SQL statement should be logged and
       a warning message should be written to the error log. However,
       such statements were instead not executed and an error message
       was written to the error log, which is the correct behavior when
       binlog_format is set to MIXED or ROW. The issue has now been
       corrected so that unsafe statements are logged with a warning as
       expected when binlog_format is set to STATEMENT. (Bug #28429993,
       Bug #73936)

     * Microsoft Windows: Validity testing for the
       named_pipe_full_access_group system variable did not account for
       NULL values. (Bug #29256690)

     * The authentication_ldap_simple plugin could enforce
       authentication incorrectly. (Bug #29637712)

     * On the Fedora 29 platform, the compat-openssl10-devel
       build dependency was changed to openssl-devel. (Bug #29278747)

     * If CMake finds a libtirpc library that is too old to work
       with MySQL, it tries to use Sun RPC from glibc instead.  (Bug
       #29240701)

     * With the --users option, mysqlpump wrote CREATE USER and
       GRANT statements to the output, but too late to apply to the
       other objects created by the dump. Consequently, restoring the
       dump file created the user accounts too late to apply to other
       objects created by the file.  mysqlpump now writes user accounts
       to the dump file before other objects. (Bug #29023216)

     * A damaged mysql.user table could cause a server exit.
       (Bug #28986737)

     * The CMake check for tirpc headers now falls back to using
       pkgconfig, to enable finding the headers on more Linux platforms.
       This requires that pkgconfig be installed.  (Bug #28970313, Bug
       #93341, Bug #28997093)

     * An attempt to access a null pointer could occur during
       prepared statement execution. (Bug #28692136)

     * The Aborted_connects status variable was not incremented
       for unsuccessful connection attempts, if connections were managed
       by the thread_pool plugin. (Bug #28490126)

     * mysqladmin shutdown did not wait for mysqld to shut down.
       (Bug #28466137, Bug #91803) References: This issue is a
       regression of: Bug #25364806.

     * Repeated invocations of stored procedures which executed
       queries undergoing short-circuit evaluation were not always
       handled correctly. (Bug #28379655)

     * Keyring migration should require only read access to the
       source keyring, but failed unless the user had write access. (Bug
       #28339014)

     * If a user performing a keyring migration did not have
       write access to the keyring file, the migration failed but
       reported success in its final error log message. (Bug #28330922)

     * During FLUSH STATUS execution, the Performance Schema
       unnecessarily aggregated session status to global status, causing
       double counts for some status variables. (Bug #28291258, Bug
       #91541)

     * Some status variable values could temporarily increase
       before returning to their original value. (Bug #27839644, Bug
       #90351)

     * Executing ALTER INSTANCE ROTATE INNODB MASTER KEY and
       migrating keys from the keyring_file plugin to the
       keyring_encrypted_file plugin could make encrypted tables
       unusable. (Bug #27760952)

     * The binary file for the udf_example user-defined function
       was omitted from binary distributions. (Bug #26115002, Bug
       #29178542)

     * When the server was started with the --skip-name-resolve
       option, spurious warnings could be written to the error log about
       ignoring accounts with a host name part of localhost. (The
       accounts in fact were used and not ignored.) (Bug #23329861, Bug
       #81441)

     * Installing and uninstalling a plugin concurrently with
       client connection activity could cause a server exit.  (Bug
       #22980441)

     * Some queries involving complex joins leaked file handles.
       (Bug #90902, Bug #28039829)

On Behalf of MySQL Release Engineering Team,
Surabhi Bhat