MySQL Server Updates & Release Notes

Note

These release notes were created with the assistance of MySQL HeatWave GenAI.

Atomic DDL Notes

• Dropping a database is now possible even if its corresponding directory is missing from the file system, maintaining a consistent state for the database. (Bug #38711166)
• Generated Invisible Primary Keys (GIPK) were added to tables even when UNIQUE NOT NULL keys existed, when sql_generate_invisible_primary_key=ON. If the unique key was autoincremented, GIPK creation failed.
  As of this release, UNIQUE NOT NULL keys are treated as Primary Key-equivalent for CREATE and ALTER statements. (Bug #38195294)
• Applying mysqlbinlog output failed while executing CREATE TABLE ... START TRANSACTION which were immediately followed by comments, preceded by #. Errors were returned similar to the following:
  ERROR 3977 (HY000) at line 117:
  Only BINLOG INSERT, COMMIT and ROLLBACK statements are allowed after CREATE TABLE with START TRANSACTION statement.
  (Bug #36313112)
• It was not possible to create tables with a Primary Key Equivalent if SQL_REQUIRE_PRIMARY_KEY was enabled. Errors were returned similar to the following:
  ERROR 3750 (HY000):
  Unable to create or change a table without a primary key, when the system variable 'sql_require_primary_key' is set. Add a primary key to the table or unset this variable to avoid this message. Note that tables without a primary key can cause performance problems in row-based replication, so please consult your DBA before changing this setting.
  (Bug #31402520)
• It was not possible to drop columns, on a table with virtual columns, using LOCK=NONE. (Bug #83557, Bug #24962142)

Audit Log Notes

• Fixed an issue related to processing certain gzip files. The MySQL Server has been updated to relax the .gz header checks, allowing ordinary gzip files to be processed. Errors were returned similar to the following:
  Invalid audit log file content
  (Bug #38980223)
• Fixed an issue related to the audit_log_prune_seconds system variable. You can now set audit_log_prune_seconds. Checks have been added to ensure audit_log_prune_seconds does not exceed a certain threshold based on other system variable values, including log_offload.log_analytics_schedule and audit_log_rotate_on_time. (Bug #38941456)
• A new feature has been introduced that rotates the audit log file based on elapsed time, controlled by the new variable audit_log.rotate_on_time. (WL #17213)
• Audit Log now supports automatic recovery from invalid filter configurations, ensuring that the server can start even if an invalid filter is found in the table. With this update, if an invalid filter configuration is detected, it will be replaced in-memory at startup with a default filter that logs all events, preventing any loss of audit activity. This feature is controlled by the global variable audit_log.filter_recovery_mode (or audit_log_filter_recovery_mode for the Audit Log plugin), which allows users to choose from three recovery modes: LOG_ALL_IF_INVALID_FILTER_DETECTED, LOG_NOTHING_IF_INVALID_FILTER_DETECTED, and ABORT_IF_INVALID_FILTER_DETECTED. This enhancement provides greater flexibility and reliability in managing audit log configurations, reducing the risk of operational issues and ensuring that critical audit data is not missed.
  See audit_log.filter_recovery_mode and audit_log_filter_recovery_mode for more information. (WL #17179)

Authentication Notes

• Authentication now supports PBKDF2 storage format with Caching_sha2_password, providing enhanced security and flexibility. This update allows for the use of PBKDF2 with SHA512, enabling a smoother migration path from existing formats and ensuring even stronger security guarantees. With this change, users can switch to the new storage format without requiring client-side modifications, and administrators can enforce the preferred storage format for added security. Overall, this enhancement strengthens password protection and offers a more robust authentication experience.
  See Caching SHA-2 Pluggable Authentication for more information. (WL #17160)

Compilation Notes

• The included zlib library has been upgraded to version 1.3.2. (Bug #38987448)

Component Notes

• The following components, previously availabile only in MySQL Enterprise Edition, are now available in MySQL Community Edition:
  • Replication Applier Metrics Component
  • Group Replication Flow Control Statistics Component
  • Group Replication Resource Manager Component
  • Group Replication Primary Election Component
    (WL #17234, WL #17235, WL #17236, WL #17237)
• The Telemetry component, previously available only in MySQL Enterprise Edition, is now available in MySQL Community Edition.
  See Telemetry. (WL #17224)

Configuration Notes

• Microsoft Windows: On Windows, MySQL Configurator did not validate group names defined with --named-pipe-full-access-group.
  As of this release, an error is returned if the group does not exist. (Bug #38068986)

Data Dictionary Notes

• When altering a table's character set, the date and time data type columns are converted in both the data dictionary and SDI. However, the new collation_id in SDI did not match that of the Field object.
  As of this release, ALTER TABLE CONVERT TO CHARACTER SET does not alter the character set for temporal types stored in data dictionary or SDI. The server always uses my_charset_latin1 for temporal types in data dictionary and SDI.
  Our thanks to Venkatesh Prasad Venugopal and the team at Percona for the contribution. (Bug #114830, Bug #36574259)

SQL Function and Operator Notes

• Fixed an issue relating to the DEFAULT() function. (Bug #39057054)

Replication with GTIDs

• A new variable, replica_allow_higher_version_source, has been introduced that lets you enable or disable replication from a higher version source into a lower version replica. (WL #17202)

InnoDB Notes

• Memory usage during FTS index construction for large tables has been optimized. (Bug #39040226)
• Fixed an issue relating to multi-value indexes. (Bug #39040128)
• Fixed an issue in the parallel reader. (Bug #39033858)
• dict_sdi_create_idx_in_mem function added columns in a different order than expected, with the compressed_len and uncompressed_len fields swapped. (Bug #38810801)
• Under certain circumstances, running CREATE INDEX with a high value for --innodb_parallel_read_threads could cause the disk space to fill up, leading to disk space exhaustion. (Bug #38370155)
• Fixed an issue related to TRUNCATE TABLE. (Bug #38169053)
• Under certain circumstances, if the cgroups memory limit was set to a very high value, and --innodb-dedicated-server=ON, MySQL could fail to start. (Bug #37944614)
• Under certain circumstances, when calculating the maximum possible index record size, an assertion failure could occur. (Bug #85060, Bug #25579578)
• MySQL Server now supports handling resource limitations set using the cpuset cgroup, providing a mechanism for assigning a set of CPUs to the application running in the cgroup. This update enables the server to adhere to the restrictions set by the cpuset-cpus cgroup controller, ensuring accurate calculation of logical CPUs available to the MySQL Server. With this enhancement, users can expect improved performance and efficient resource utilization, as the server can now correctly determine the number of logical CPUs according to cgroup restrictions. (WL #16851)
• MySQL Clone plugin now supports cloning between consecutive LTS versions higher than 9.7.0.
  See Clone Plugin Limitations and Remote Cloning Prerequisites. (WL #16520)

JSON Notes

• Fixed an issue relating to processing of JSON files. (Bug #39000847)
• Using the monthname function within a JSON scope resulted in an assertion error. Errors were returned similar to the following:
  Assertion `mtime.time_type == MYSQL_TIMESTAMP_DATE in JSON
  (Bug #38980997)
• MySQL Community Server now supports DML operations on JSON Duality Views, enabling users to perform insert, update, and delete operations on these views. This enhancement expands the functionality of JSON Duality Views, which were previously limited to DDL operations in the Community Server edition. With this update, users can now fully leverage the capabilities of JSON Duality Views in MySQL Community Server, making it easier to manage and manipulate data in a flexible and efficient manner.
  See DML Operations on JSON Duality Views for more information. (WL #17246)
• MySQL now supports the use of auto-increment columns in DML operations with JSON duality views, allowing for more flexible and efficient data management. With this update, you can leverage auto-generated values for primary keys, enabling easier insertion of data into JSON duality views. The system automatically handles the generation of these values, streamlining the data insertion process and reducing the need for manual intervention. This enhancement provides a more seamless and intuitive experience for users working with JSON duality views, making it easier to manage and manipulate data in a variety of scenarios. (WL #17056)

Optimizer Notes

• Fixed an issue relating to optimized prepared DELETE and UPDATE statements. (Bug #39071552)
• Fixed an issue relating to Batched Key Access path and partitioning. (Bug #38947039)
• Fixed an issue relating to query management. (Bug #38928287)
• Fixed an issue relating to memory management. (Bug #38573278)
• TIMEDIFF() returned the wrong result if the first argument was DATETIME and the second argument was DATE. (Bug #38181443, WL #16895)
• TIMEDIFF() did not return NULL if unsupported values were used. (Bug #38179658, WL #16895)
• FROM_DAYS() did not handle out of range values, or values less than 366, consistently. (Bug #38177844, Bug #38177993, WL #16895)
• DAYNAME() did not return the day name if used in an arithmetic expression but returned the day number instead. (Bug #38177821, WL #16895)
• ADDDATE() returned an incorrect result if the first argument was year zero (0000). (Bug #38177766, WL #16895)
• When copying a DATE value to an INTEGER column, the DATE value was extended to DATETIME and 00:00:00 was added to it. (Bug #38043311)
• Fixed an issue relating to the Optimizer's SQL planner. (Bug #35634700)
• Overflow handling of date arithmetic has been improved. (Bug #32019977)
• CAST() of a year column returned the wrong result. (Bug #29616536, Bug #35093107, WL #16895)
• Under certain circumstances, with ALLOW_INVALID_DATES sql_mode enabled, invalid dates were not properly handled and no warning was issued if they were set to NULL. (Bug #20583945, WL #16895)
• ADDTIME, SUBTIME, ADDDATE, and SUBDATE did not return NULL for invalid values. (Bug #12805124)
• Under certain circumstances, find_in_set() could return incorrect results. The optimization which caused this has been removed.
  Our thanks to Jingqi Tian for the contribution. (Bug #119995, Bug #39056621)
• When errors and warnings occured during background histogram updates, the conditions in the diagnostic area were not cleared, leading to future background histogram updates emitted the same stale conditions to the error log.
  Our thanks to Tony Chen and the team at Amazon for the contribution. (Bug #119922, Bug #38983545)
• IN operator returned an unexpected result if one of values in the list is NULL. (Bug #119650, Bug #38831745)
• Datetime comparisons were inconsistent depending on the type used to store the values, INT or BIGINT. (Bug #119136, Bug #38819341, WL #16895)
• Fixed an issue with NULL detection for temporal data types in derived tables after a derived merge.
  Our thanks to Xingyu Yang and the team at Tencent for the contribution. (Bug #118658, Bug #38191248)
• Under certain circumstances, incorrect results were returned when comparing DECIMAL columns to constants containing more decimals than the column. (Bug #118033, Bug #37864937)
  References: This issue is a regression of: Bug #34361287.
• A previous fix changed range analysis for non-binary string collations, when a string constant in a range predicate was longer than the indexed column, no index range was formed, unless the collation was known to be safe. This resulted in queries, which should have used efficient range plans, using index or table scans.
  Our thanks to Yakir Gibraltar for the contribution. (Bug #118009, Bug #37849917)
  References: This issue is a regression of: Bug #35169384.
• Under certain circumstances, INTERSECT could return an incorrect result. (Bug #117911, Bug #37804715)
• A histogram generated by ANALYZE TABLE UPDATE HISTOGRAM ON col_name was not be accepted by ANALYZE TABLE UPDATE HISTOGRAM ON col_name USING DATA ... if the histogram was built on a bigint column with values outside the uint32 range.
  Our thanks to Tianfeng Li for the contribution. (Bug #116611, Bug #37269033)
• In an explain expanded query, join order hints were not printed with a valid syntax.
  Our thanks to Kaiwang Chen for the contribution. (Bug #116084, Bug #37053765)
• Conditions on const tables and on previously joined tables were not consistently taken into account when building ranges in get_quick_record_count(). This could lead to suboptimal or incorrect range estimates, especially when range predicates depended on values from const or earlier-joined tables.
  Our thanks to Xingyu Yang for the contribution. (Bug #112737, Bug #35912840)
• Using DISTINCT with date functions returned inconsistent results. (Bug #109351, Bug #34889757)
• Assigning a date value to a bigint column in MySQL resulted in zeroes being appended to the end of the value, unless an explicit cast was used. Support for automatic casting in assignment operations has been added when the right-hand side is a single value or column. (Bug #79563, Bug #22353325, WL #16895)
• The Hypergraph Optimizer is now available in MySQL Community Edition.
  It can be enabled in one of the following ways:
  • Session scope: SET optimizer_switch='hypergraph_optimizer=on|off'
  • Global scope: SET GLOBAL optimizer_switch='hypergraph_optimizer=on|off'
  • Persisted scope: SET PERSIST optimizer_switch='hypergraph_optimizer=on|off'
  • Server startup: mysqld --optimizer-switch='hypergraph_optimizer=on|off'
  • Per-statement hint: /*+ SET_VAR(optimizer_switch='hypergraph_optimizer=on|off') */
    (WL #17265)

Packaging Notes

• Building with PGO (Profile Guided Optimization) is now also supported for RPM builds on SLE/openSUSE and Fedora, by adding --define=with_pgo 1 to the rpmbuild command line. (Bug #38915967)
• For platforms on which OpenSSL libraries are bundled, the linked OpenSSL library for MySQL Server has been updated to version 3.5.5. For more information, see OpenSSL 3.5 Series Release Notes. (Bug #38867109, WL #17209)

Performance Schema Notes

• Prepared statements with mismatched parameter types could enter infinite loops and consume 100% CPU after re-preparation.
  Our thanks to Alex Xing for the contribution. (Bug #38600714)

Thread Pool Notes

• Under high load, when using Thread Pool, connections could hang. (Bug #38693319)

Functionality Added or Changed

• X Protocol now supports PBKDF2 storage format with Caching_sha2_password. (Bug #38879953)

Bugs Fixed

• It was not possible to use double quotes in revoke_schema_privileges_from_all_accounts_except if ANSI_QUOTES was present in sql_mode. (Bug #39105598)
• Fixed an issue relating to Information Schema privileges. (Bug #35243416, Bug #37387633)
• The SETS keyword was not included in the parser's identifier keyword lists, preventing its use as an unquoted identifier. (Bug #119904, Bug #38974036)
• The SQL parser consumed a large amount of memory when parsing very large queries with many large IN clauses.
  As of this release, the SQL parser's memory management has been improved. (Bug #105004, Bug #33390851)

Original source

Note

These release notes were created with the assistance of MySQL HeatWave GenAI.

Audit Log Notes

• Updated the audit_log_rotate_on_size system variable to require the AUDIT_ADMIN privilege for modification. (Bug #38620615)
• MySQL now supports a more modular and efficient auditing system with the introduction of the Audit Log component. This update splits the monolithic Audit Log software into smaller, more manageable components, encapsulated into MySQL components, making it easier to install, manage, and maintain. The new component-based architecture allows for improved flexibility and customization, enabling users to easily configure audit log settings, such as output file location, format, and buffer size. With this enhancement, MySQL provides a more streamlined and user-friendly auditing experience, allowing you to focus on monitoring and analyzing your database activities with greater ease.

See MySQL Enterprise Audit Component for more information. (WL #12716)

Compilation Notes

• MySQL Server's clang-tidy configuration now consistently disables all brace checks, including hicpp-braces-around-statements, to prevent duplicate diagnostics. (Bug #38601902)
• A number of unused symbols were removed from the code base. (Bug #38557463)
• The cmake macro MYSQL_CHECK_PROTOBUF did not properly handle multiple calls when using system protobuf. Errors were returned similar to the following:

ADD_LIBRARY cannot create imported target "ext::libprotobuf" because another target with the same name already exists.

(Bug #38417986)

Configuration Notes

• The mysqlx_port server variable was not persisted in the my.ini file when set through the Type and Networking configuration page. (Bug #38728671)
• MySQL Configurator treated single-quotes (') as part of the password. The configurator now correctly handles quoted passwords by removing any quotes, double quotes, or backticks from the beginning and end of the password. (Bug #38069029)
• Server file permissions were not applied as specified with --server-file-permissions-access. Full access was granted regardless of the value defined. (Bug #38068994)
• --slow-query-log-file could be used even if the slow query log was not enabled. (Bug #38068822)
• It was possible to set the --general-log-file CLI option when the general log was disabled. As of this release, an error is returned. (Bug #38068775)
• --open-win-firewall could be used when TCP/IP connections were disabled. As of this release, --open-win-firewall can only be used when TCP/IP connections are enabled. (Bug #38068699)
• Updating the password with MySQL Configurator took a long time. (Bug #37459271)

Doxygen Notes

• A number of issues were corrected in the Server API documentation. Our thanks to Daniël van Eeden for the contribution. (Bug #114671, Bug #117516, Bug #37329988, Bug #37615382, Bug #36526371)

InnoDB Notes

• InnoDB redo logging error messages now include the current LSN and redo log capacity information. A new warning code ER_IB_WRN_REDO_DISABLED_INFO is added which includes the current LSN. A new error code ER_IB_MSG_LOG_WRITER_WAIT_ON_NEW_LOG_FILE_INFO is added to include the current log capacity and the current log capacity used. The MONITOR output has been enhanced to include redo log capacity details and current log capacity.
• The warning ER_IB_WRN_REDO_DISABLED was removed, as was the error ER_IB_MSG_LOG_WRITER_WAIT_ON_NEW_LOG_FILE. (Bug #37645185)
• If an XA PREPARED transaction existed during a server shutdown or crash, restarting the server with a non-zero --innodb-force-recovery value caused InnoDB to incorrectly change the transaction state from PREPARED to ACTIVE. This could lead to issues ranging from assertion failures to rolling back the PREPARED transaction. (Bug #35659774)
• InnoDB now generates unique rowid values in tables without primary keys more efficiently. (Bug #13395168)
• The field fk_max_recusive_level in dict_table_t struct is no longer used and has been removed. Our thanks to Xiaocong Ding for the contribution. (Bug #119063, Bug #38492574)
• If the server closed during transaction commit, transactions in an already committed state could remain in the undo log. If the transaction is already committed, it is cleared from trx_sys->rw_trx_list during recovery, but is not cleared from active_tw_trxs of the trx_sys->shards. This invalid transaction stored in the trx_sys->shards could cause errors or assertions in the execution of the later queries. As of this release, during cleanup on startup, processed transactions are not added to trx_sys->shards. Our thanks to Yewei Xu and the team at Tencent for the contribution. (Bug #118515, Bug #38132549)
• When selecting blob data after a small partial update, incorrect results may be returned due to inconsistencies in the Multi-Version Concurrency Control (MVCC) for blob objects. (Bug #114138, Bug #36342348)
• Concurrent execution of FLUSH TABLE FOR EXPORT with DROP TABLE and DML operations could result in a deadlock. (Bug #110485, Bug #35221441)
• Fixed an issue related to bulk inserts. (Bug #38208188)
• MySQL Server now supports a new startup option, container_aware, which controls the server's ability to discover and adhere to CPU and memory resource limits set by a container. (WL #16937)

JavaScript Programs

• GraalVM performance metrics are now available on Windows. (Bug #38498718)
• The SHOW CREATE LIBRARY command generated HEX representations of binary libraries on x86_64 architectures with FFFFFF patterns before each byte with the high bit set. Our thanks to Akshat Nehra and the team at Amazon for the contribution. (Bug #118927, Bug #38458235)

JSON Duality Views

• The table IDs reported in the I_S views for JSON Duality Views differed from those reported by ORACLE DB due to differences in calculation order, leading to inconsistent REFERENCED_TABLE_ID values when multiple Content_tree_nodes referred to the same base table. (Bug #38624596)

Logging Notes

• MySQL now supports enhanced debugging capabilities with the addition of timestamps to GCS/XCOM trace file entries. For more information, see the group_replication_communication_debug_options system variable documentation in Group Replication System Variables. (WL #17008)

MySQL Enterprise Notes

• Option Tracker now supports MySQL Shell and MySQL Shell for VS Code.
• The following status variables were added:
  • option_tracker_usage-MySQL_Shell_VSC_Natural_Language_to_SQL
  • option_tracker_usage-MySQL_Shell_VSC_HeatWave_Chat
  • option_tracker_usage-MySQL_Shell_VSC_Natural_Language_to_SQL
  • option_tracker_usage-MySQL_Shell_VSC_Lakehouse_Navigator
  • option_tracker_usage-MySQL_Shell
  • option_tracker_usage-MySQL_Shell_for_VS_Code
  • option_tracker_usage-MySQL_Shell_Dump
  • option_tracker_usage-MySQL_Shell_VSC_Dump
  • option_tracker_usage-MySQL_Shell_Dump_Load
  • option_tracker_usage-MySQL_Shell_VSC_Dump_Load
  • option_tracker_usage-MySQL_Shell_MRS
  • option_tracker_usage-MySQL_Shell_VSC_MRS
  • option_tracker_usage-MySQL_Shell_Copy
  • option_tracker_usage-MySQL_Shell_Upgrade_Checker
• The following system variable was added:
  • component_option_tracker.mysql_shell_support

See Option Tracker Component. (WL #16659, WL #17000)

Optimizer Notes

• Fixed an issue relating to Common Table Expressions (CTE). (Bug #38573285)
• Using the coalesce or any_value functions with specific JSON operations, such as json_search and json_remove, could cause an assertion failure. (Bug #38549573)
• Negative year values were accepted as valid datetime values by one method, but not by another. Negative year values are now always reported as errors. (Bug #38487373)
• Fixed issues relating to executing specific SELECT queries. (Bug #38465147, Bug #38465178)
• Fixed an issue related to executing specific SQL queries. (Bug #38448700)
• Fixed an issue relating to SHOW CREATE TABLE. (Bug #38298692)
• As of this release, GROUPING() is permitted in queries without ROLLUP. (Bug #38168051)
• Queries which used REGEXP took longer to execute as prepared statements than as direct queries. (Bug #114056, Bug #36326728)

Packaging Notes

• For platforms on which OpenSSL libraries are bundled, the linked OpenSSL library for MySQL Server has been updated to version 3.0.18. For more information, see OpenSSL 3.0 Series Release Notes. (Bug #38632932)

Performance Schema Notes

• Important Change: It is now possible to view accounts which have been temporarily locked, using the new Performance Schema table TEMPORARY_ACCOUNT_LOCKS.
• The Performance Schema table, HOST_CACHE, was extended with the following columns:
  • COUNT_ACCOUNT_LOCKED_ERRORS: The number of errors caused by a permanently locked account.
  • COUNT_TEMPORARY_ACCOUNT_LOCKED_ERRORS: The number of errors caused by a temporarily locked account.
    (Bug #37378977, WL #16803)
• Fixed an issue relating to querying variables. (Bug #38100668)
• Performance Schema did not capture SQL or digests for prepared statements from non-MySQL clients. (Bug #114480, Bug #36445818)
• Negative numbers on digests in WHERE were not parsed properly. Our thanks to Yakir Gibraltar for the contribution. (Bug #110045, Bug #35078042)
• The following logs are now instrumented for Telemetry Logging:
  • Slow Query Log
  • General Log

See The setup_loggers Table and Configuring Log Telemetry. (WL #17167)

Pluggable Authentication

• Connecting to a MySQL server with a non-existing user now consistently returns an "Access denied for user" error message, regardless of the username length or MySQL version. Previously, errors were returned similar to the following:

ERROR 1045 (28000): Access denied for user 'foo'@'localhost' (using password: NO), ERROR 1524 (HY000): Plugin 'mysql_native_password' is not loaded

(Bug #36527984)

Replication Notes

• MySQL Server now supports a new GTID set data structure, providing a simpler, more modern, and efficient library for handling Global Transaction IDs. This implementation replaces the existing library, offering improved maintainability and performance. (WL #16076)

Security Notes

• Important Change: MySQL now supports enhanced security and flexibility with the relocation of MD5() and SHA1() SQL functions to a separate component, allowing for greater control over deprecated hashing algorithms and improved compliance with security standards. You can install the classic_hashing component to continue using MD5() and SHA1() functions in your applications, ensuring a seamless transition and maintaining the integrity of your database operations. This update enables you to choose whether to utilize these functions, promoting a more secure environment by avoiding unacceptable hashing algorithms as defined by industry standards.

See Legacy Hashing Component for more information. (WL #16956)

• Fixed an issue relating to authentication. (Bug #118447, Bug #38077617)
• The error message returned when attempting to create a user with an existing user name was unclear. (Bug #28331, Bug #11746738)

SQL Syntax Notes

• MySQL Server now supports table-level DML tagging for JSON duality views. You can explicitly specify allowed DML operations (INSERT, UPDATE, DELETE) per table when defining a JSON duality view. Restrictive tags (NO INSERT, NO UPDATE, NO DELETE) are also supported to explicitly disallow specific operations. You can specify combinations of these tags. At runtime, JSON duality views validate incoming DML operations against the configured tags and accept or reject operations accordingly. This provides per-table control over write operations within a single duality view definition. The behavior is compatible with Oracle DB JSON duality views.

See CREATE JSON DUALITY VIEW Statement for more information. (WL #17053)

Functionality Added or Changed

• Important Change: In previous releases, Foreign Key constraints and cascade operations were handled by the InnoDB storage engine. As of this release, they are handled by the SQL layer of the MySQL Server. This change allows for more comprehensive binary logs and improved analytics. This change is fully backward compatible and ensures that data changes are always visible, logged, and replicated.

To continue using InnoDB to handle Foreign Key constraints and cascade operations, you must start the server with the new system variable, innodb_native_foreign_keys. (WL #11249)

• The bundled version of opentelemetry-cpp was upgraded to version 1.23.0. (Bug #38483354)

Bugs Fixed

• InnoDB: Under certain circumstances, when using the que_eval_sql interface, a race condition could occur. (Bug #118705, Bug #38310595)
• mysqltest did not print the symbolic error code (error name) for failed SQL statements. (Bug #38743161)
• Running SET PERSIST on a system variable after an upgrade can result in duplicate variable entries across different sections in mysqld-auto.cnf. (Bug #38680162)
• (Bug #38560000)
• Binary log purged before persisted binlog_expire_* options were loaded. (Bug #38554467)
• Fixed an issue relating to changing users and roles. (Bug #38492547)
• Linux debug builds started with the --basedir option set to an existing absolute path will no longer assert that local_mysql_home ends with a specific character. (Bug #38483568)
• Releasing MDLs before the Auto_releaser has run could cause asserts in debug builds due to locks not being released. (Bug #38430539)
• When 2 of 6 instances in a geographically dispersed InnoDB Cluster lost connectivity, the primary server became unresponsive, causing the COMMIT and INSERT operations to become unresponsive as well. (Bug #38380392)
• Fixed an issue related to user name handling. (Bug #38355483)
• mysqldump's --order-by-primary option sorted data by every index on the table, instead of just sorting by the primary key. (Bug #38284832)
• Fixed several issues relating to Thread Pool blocking connections from closing. (Bug #38170188, Bug #36782728, Bug #38549372)
• Under certain circumstances, when used for an upgrade, mysqlpump altered the SQL mode and collation of stored procedures. (Bug #38154661)
• Fixed several issues in the parser. (Bug #37866044, Bug #38483735)
• Upgraded the MySQL Server to use non-deprecated OpenSSL algorithms for enhanced security and compatibility. (Bug #33951773)
• It was possible to specify a password expiry policy for externally authenticated users. (Bug #32724080)
• GTID gaps were generated because of the replica-skip-errors option. (Bug #28590993)
• Database names containing a backslash were not correctly quoted when using the --routines option with mysqldump. (Bug #22240504)
• If server was started on Windows platforms with --skip-grant-tables, privileges were not checked and remote connections were not permitted. As a result, the server might not start. (Bug #118805, Bug #38328780)
• (Bug #97091, Bug #30381098)
• Wait_timeout and the sql_mode MODE_IGNORE_SPACE were set at connect time, only. They should also have been set at reset_connection time. Our thanks to Facebook for the contribution. (Bug #97091, Bug #30381098, Bug #97666, Bug #30551153)
• A memory management issue occurred in the CSV engine when meta files became inaccessible. Our thanks to Yan Huang for the contribution. (Bug #93394, Bug #28987279)
• Trailing comments on the !includedir or !include directives in the MySQL configuration file could prevent the server from starting. (Bug #87657, Bug #26750972)

Original source

Note

These release notes were created with the assistance of MySQL HeatWave GenAI.

Account Management Notes

• Account Management Notes
• Audit Log Notes
• Authentication Notes
• Binary Logging
• Compilation Notes
• Component Notes
• Configuration Notes
• Connection Management Notes
• Deprecation and Removal Notes
• SQL Function and Operator Notes
• InnoDB Notes
• JavaScript Programs
• JSON Notes
• JSON Duality Views
• Keyring Notes
• MySQL Programs
• Optimizer Notes
• Packaging Notes
• Performance Schema Notes
• Pluggable Authentication
• Replication Notes
• Spatial Data Support
• sys Schema Notes
• Thread Pool Notes
• Bugs Fixed

Account Management Notes

• MySQL now supports enhanced role management with the introduction of a new option, activate_mandatory_roles.
  If activate_all_roles_on_login is enabled, activate_mandatory_roles is ignored, and both mandatory and granted roles are activated. If activate_all_roles_on_login is disabled and activate_mandatory_roles is enabled, mandatory roles are activated in addition to any default roles associated with the account.
  If neither are activated, only the account's default roles are activated.
  activate_mandatory_roles is enabled by default. (WL #16597)

Audit Log Notes

• Fixed an issue related to audit log strategy. (Bug #38223011)
• Improved performance related to adding multiple general_sql_command.str fields to an Enterprise Audit filter. (Bug #38053242)
• Fixed an issue related to audit_log_read performance. (Bug #36281295, Bug #38275868)

Authentication Notes

• Security Enhancement: The default value of caching_sha2_password_digest_rounds is increased to 10,000. (WL #16974)
• Under certain circumstances, the keyring component could not read the configuration file if the configured data directory path was shorter than the default path. (Bug #38078884)
• General-Purpose Keyring Functions did not handle invalid parameters correctly. (Bug #38055488)

Binary Logging

• Important Change: The default value for the binlog_transaction_dependency_history_size server system variable has been increased from 25000 to 1000000 (one million). This variable's maximum value has also been increased, from 1000000 to 10000000 (ten million).
  This change should not have any immediate effect on existing setups.
  See the description of binlog_transaction_dependency_history_size for further information. (WL #16439)

Compilation Notes

• Solaris: Clang and GCC now can be used for compiling MySQL on Solaris. (Bug #30562248)
• OpenSSL 3 is now used on Solaris. (Bug #38193362)
• Upgraded the bundled googletest library to version 1.17. (Bug #38152017)
• Upgraded the bundled libbacktrace library to the latest version (April 2025). (Bug #38151970)
• MySQL Server now bundles libzip and provides CMAKE options to choose between the bundled or system library when compiling the server. (Bug #38109824)
• The minimum required version of CMAKE is now 3.17.5, for supported platforms. (Bug #38090672)
• Compiler warnings were logged when building MySQL Server with GCC 15. (Bug #37721464)

Component Notes

• The MYSQL_OPTION.OPTION_USAGE.OPTION_NAME field used a different character set than PERFORMANCE_SCHEMA.MYSQL_OPTION.OPTION_NAME.
  As of this release, the CLUSTER_ID, SERVER_ID, and OPTION_NAME fields of MYSQL_OPTION.OPTION_USAGE use utf8mb4_general_ci instead of ascii. (Bug #38230424)
• The sql_errno() method in the mysql_command_error_info service returned 0 when a command failed, instead of the appropriate error code. (Bug #38178087)
• The server plugin system variable flag PLUGIN_VAR_NOSYSVAR was not processed properly. Variables defined with this flag were displayed when they should be hidden. (Bug #37654916)
• Errors were not handled correctly when using the keyring UDF plugin and executing specific queries with the keyring_key_generate function. (Bug #37501458)
• Telemetry component introduces a secret decoding service. This enables you to create your own secret-decoding service for OpenTelemetry.
  The following system and status variables were added:
  • telemetry.otel_exporter_otlp_traces_secret_headers
  • telemetry.otel_exporter_otlp_metrics_secret_headers
  • telemetry.otel_exporter_otlp_logs_secret_headers
  • telemetry.resource_provider
  • telemetry.secret_provider
  • telemetry.run_level
    (WL #16647)

Configuration Notes

• Microsoft Windows: If the MySQL Server was configured as a process, not as a service, the Configurator's Starting the server step could take a long time to complete due to delays in detecting when the server was ready to receive connections. (Bug #38068649)
• Microsoft Windows: Shared Memory connections, --enable-shared-memory=true, failed when they were the only enabled protocol during server configuration. (Bug #38068619)
• Microsoft Windows: It was not possible to create usernames with escaped characters, using single or double quotes, during new server configuration. (Bug #37459624)
• Microsoft Windows: When upgrading, for example from version 9.1.0 to 9.2.0, using a non-default port, such as 3307, messages could be generated indicating a connection issue to port 3306, even if the server configuration is correct and the upgrade is successful. Errors were returned similar to the following:
  Can 't connect to MySQL server on 'localhost': 3306' (10061)
  (Bug #37459255)
• Microsoft Windows: The CLI help for MySQL Configurator now correctly displays the list of supported values for configurable server options. (Bug #37458648)
• --config-type Configurator CLI option did not accept any values. (Bug #38068684)

Connection Management Notes

• Connection control now supports exempting unknown users from delay criteria, introducing a new system variable component_connection_control.exempt_unknown_users that allows administrators to configure whether failed connection attempts without user credentials should be penalized. This enhancement improves the component's ability to handle legitimate connection attempts from load balancers, ensuring better server availability while maintaining effectiveness in thwarting brute force attacks. A new status variable component_connection_control_exempted_unknown_users provides insights into the number of exempted connections.
  For more information, see (xref linkend="to-fill"). (WL #17037)

Deprecation and Removal Notes

• The group_replication_allow_local_lower_version_join system variable, deprecated in MySQL 8.4.0, has now been removed. (WL #16019)
• The replica_parallel_type system variable, deprecated in MySQL 8.0.29, has now been removed. (WL #13955)
• The SCRAM-SHA-1 authentication method for SASL LDAP authentication is now deprecated, and subject to removal in a future version of MySQL. You are encouraged to use SCRAM-SHA-256 instead.
  As part of this work, the default value for the authentication_ldap_sasl_auth_method_name system variable has been changed to SCRAM-SHA-256; the value SCRAM-SHA-1 is now deprecated.
  For further information, see LDAP Pluggable Authentication. (WL #17022, WL #17034)
• The following deprecated plugins are removed in this release:
  • semisync_master replaced by semisync_source.
  • semisync_slave replaced by semisync_replica.
    (WL #16801)

SQL Function and Operator Notes

• Fixed an issue related to executing certain UPDATE statements. (Bug #37590580)
  References: This issue is a regression of: Bug #31562881.

InnoDB Notes

• Important Change: The default value for innodb_log_writer_threads is now determined in part by whether binary logging is enabled, as shown here:

  if (log_bin = OFF) {
    if ([number of logical CPUs] <= 4) {
      innodb_log_writer_threads = OFF
    } else {
      innodb_log_writer_threads = ON
    }
  } else {
    if ([number of logical CPUs] <= 32) {
      innodb_log_writer_threads = OFF
    } else {
      innodb_log_writer_threads = ON
    }
  }
  

  When binary logging is disabled (log_bin is OFF), the default is determined as it was in MySQL 9.4 and earlier.
  This change does not affect the variable's configured value, if set.
  For further information, see the description of innodb_log_writer_threads in the MySQL documentation; see also Optimizing InnoDB Redo Logging. (WL #16999)
• InnoDB: The dict_stats_analyze_index() function could discard optimal index-based access paths, resulting in suboptimal performance for large tables when empty index statistics are read concurrently or an error leads to empty index statistics.
  Our thanks to Casa Zang and the Tencent team for the contribution. (Bug #33472935)
• Creating a thread for parallel scan may fail, resulting in assertion failures when it fell back to single-thread mode. Errors were returned similar to the following:
  Assertion failure: row0pread.h:306: active >= n_threads
  (Bug #38325137)
• Under certain circumstances, a virtual index rollback could fail on 32-bit builds of MySQL Server. (Bug #38167527)
• The mlog read_1_bytes comment now accurately reflects a byte size of 1 byte, rather than 2 bytes.
  Our thanks to Jeong Geun Lee for the contribution. (Bug #38001828)
• If the buffer pool was very large, it could lead to a large number of chunks required by each buffer pool instance, which could fail if the operating system was unable to allocate the required memory.
  As of this release, this allocation is checked to ensure it is allocated correctly. (Bug #37994397)
• Users could encounter an assertion failure due to a data size and bounds mismatch. Errors were returned similar to the following:
  Assertion failure: ddl0file-reader.cc:193: m_ptr + data_size < m_bounds.
  second
  (Bug #37882398)
  References: This issue is a regression of: Bug #37233273.
• Fixed an issue relating to modifying the internal FTS configuration. (Bug #37792010)
• Fixed an issue related to the purge coordinator.
  Our thanks to Yin Peng and the Tencent team for the contribution. (Bug #37628911)
• Fixed an issue related to virtual indexes. (Bug #37602657)
• MySQL Server could behave unpredictably when simultaneously optimizing and updating a table. (Bug #37324137)
• If the row size exceeded the maximum allowed size due to innodb_strict_mode=OFF, confusing warning messages similar to "Cannot add field" could appear in certain situations, such as selecting from a table or dropping a column.
  Error messages are improved for this scenario. (Bug #37003342, Bug #36768046, Bug #36867372)
• The dict_stats_analyze_index() function in could discard optimal index-based access paths, resulting in suboptimal performance for large tables when empty index statistics are read concurrently or an error return leads to empty index statistics.
  Our thanks to Casa Zang and the Tencent team for the contribution. (Bug #33472935)
• Under certain, rare circumstances, a race condition was possible during a purge operation. (Bug #32679024)
• As of this release, the default value of innodb_change_buffering is ALL. This improves handling of secondary index updates. (WL #16967)

JavaScript Programs

• Size limit for result sets in SQL-callouts from JavaScript stored procedures has been increased to support larger data sizes. (Bug #38413760)
• JavaScript language support provided by the MySQL Multilingual Engine component now conforms to the ECMAScript 2025 Language Specification (ECMA-262, 15th Edition) as shown at https://262.ecma-international.org/16.0/. (Bug #38358516)
• Virtual FileSystem paths were not handled correctly by the MLE component. As of this release, paths with redundant or relative directories, such as ./ or ../, are accepted without errors. (Bug #38198011)
• MySQL now supports automatic configuration for the mle.memory_max variable setting on-premises, utilizing a formula-based approach to dynamically allocate memory. This ensures that the default value of this variable is calculated as 5% of the total physical memory, capped within a range of 0.4GB to 32GB.
  For more information, see MLE Component Memory and Thread Usage. (WL #17033)

JSON Notes

• Parsing very-large, high-precision float numbers in exponential notation, such as 5.0000000000000004e36, in a JSON column could result in a checksum verification error. (Bug #38033684)
• Large TIME values inserted into a TIME column became smaller when converted to JSON. (Bug #37458212)
• Fixed an issue related to handling JSON key values in the mysql.user.user_attribute column when using the ALTER USER statement with a JSON attribute containing a blank or null key-value pair. (Bug #31067575)

JSON Duality Views

• Using 0 as an explicit request for AUTO_INCREMENT columns was allowed. (Bug #38319365)
• It was possible to delete a singleton child of a JSON Duality View's nested child. As of this release, it is no longer possible to delete a singleton child of a JSON Duality View's nested child. (Bug #38305543)
• Creating a JSON Duality View with WHERE conditions that contained sub-queries resulted in an assertion failure, and the server possibly closing unexpectedly. Errors were returned similar to the following:
  Assertion `nullptr != dynamic_cast (arg) ' failed.
  (Bug #38187473)
• INSERT statements for a JSON Duality View that included an AUTO_INCREMENT column were rejected even if the column was not part of a join condition. (Bug #38179925)
• JSON Duality Views exhibited inconsistent behavior with DML queries and SELECT queries when the derived_merge=off optimizer hint was used. (Bug #38175557)
• Executing a JSON Duality Views UPDATE statement could result in the server closing unexpectedly. Errors were returned similar to the following:
  Address not mapped to object
  (Bug #38084877)

Keyring Notes

• component_keyring_kmip or keyring_okv now support SECRET key types on Oracle Key Vault (OKV) server 21.10, or higher. (WL #17090)
• The keyring_okv plugin is deprecated and replaced by the keyring_kmip component, enabling secure communication with Key Management Interoperability Protocol (KMIP) 1.1-compatible products such as Oracle Key Vault, Gemalto SafeNet KeySecure Appliance, Townsend Alliance Key Manager, and Entrust KeyControl. (WL #16524)

MySQL Programs

• This version includes the MySQL Diagnostic Monitor (mysqldm), a new client tool designed to collect server diagnostic information. To supply to Oracle Support for efficient issue resolution, for example. The mysqldm client tool runs a predefined set of queries on the server, storing the results in JSON format, and provides options for customizing the output directory, number of iterations, and delay between iterations.
  Note
  MySQL Diagnostic Monitor is only available with MySQL Enterprise Edition.
  (WL #16973)

Optimizer Notes

• Sorting with row IDs could produce incorrect results when the result came from a hash join reading a table in the temptable engine, where the temporary table is large enough to be stored on disk. (Bug #38418831)
• Queries using semi join with materialization may return incorrect results due to a missing condition.
  Our thanks to Jingqi Tian and the team at Alibaba. (Bug #38110792)
• Fixed an issue related to query execution. (Bug #37847144)
• Separate evaluation functions for DATE and DATETIME values have been introduced, utilizing distinct classes for these data types to optimize server performance. (Bug #37743288)
• Fixed an issue relating to SQL queries involving correlated scalar subqueries and JSON tables. (Bug #36684370)
• Queries using prepared statements to retrieve time type data did not always return expected results, regardless of whether precision was specified. (Bug #36065557)
• Using WHERE IN (NULL,'xx:xx:xx.xxxxxx'); on a TIME column returned an empty set, whereas using WHERE IN ('xx:xx:xx.xxxxxx') returned the expected results. (Bug #35500028)
• Using the IN() operator with TIME columns could return an empty set when a value is matched if the list contains NULL or empty strings. (Bug #34188599)
• The default values of the following system variables have changed:
  • explain_format: changed from TRADITIONAL to TREE
  • explain_json_format_version: changed from 1 to 2.
    (WL #17039)

Packaging Notes

• For platforms on which OpenSSL libraries are bundled, the linked OpenSSL library for MySQL Server has been updated to version 3.0.17. For more information, see OpenSSL 3.0 Series Release Notes. (Bug #38457056)
• MySQL Server RPM installation on Fedora 42 could fail due to conflicts with MariaDB 11.8 packages. Errors were returned similar to the following:
  file /usr/bin/mysqld conflicts between mysql-community-server-8.4.7 and mariadb11.8-server-3.8.3, file /usr/share/man/man1/mysql.1.gz conflicts between mysql-community-client-8.4.7 and mariadb11.8, file /usr/lib/systemd/system/mysqld.service conflicts between mysql-community-server-8.4.7 and mariadb11.8-server
  (Bug #38384844)
• The RPM specification and DEB control files were missing several requirements, making it impossible to build those formats. (Bug #38310603)
• MySQL Server now supports side-by-side installations of different innovation and LTS releases. (Bug #38104701)

Performance Schema Notes

• Invalid temporal values in the binary protocol were silently adjusted when SQL_MODE was set to a non-strict value. As of this release, such values are rejected with an error. (Bug #38352728)
• The following telemetry system variables were enabled by default, resulting in connection failures and log spamming:
  • telemetry.log_enabled
  • telemetry.metrics_enabled
  • telemetry.trace_enabled
    As of this release, these variables are disabled by default. (Bug #38345446, Bug #38347061)
• The default values for the following telemetry endpoint system variables, such as http://localhost:4318/v1/traces, caused log spamming and were defined using http instead of https:
  • telemetry.otel_exporter_otlp_logs_endpoint
  • telemetry.otel_exporter_otlp_metrics_endpoint
  • telemetry.otel_exporter_otlp_traces_endpoint
    As of this release, these variables have no default value. (Bug #38133218)
• Connections waiting on the connection control plugin were not displayed in the performance_schema.processlist table. (Bug #38043683)
• Queries with the same SQL_DIGEST could produce identical summaries, even if they had distinct characteristics, such as differing ORDER BY clauses. The digest computation has been improved to better distinguish between these queries. (Bug #31103386)

Pluggable Authentication

• Client-server handshake errors occured during authentication with the authentication_oci plugin, when using the --default-auth parameter, if the server and client's default authentication methods were different on first connection. Errors were returned similar to the following:
  An error occurred during the client server handshake.
  (Bug #37092159)

Replication Notes

• MySQL Replication supports enhanced security features with encryption now being enabled for all replication connections by default. With this configuration change, the default value of the SOURCE_SSL option in the CHANGE REPLICATION SOURCE TO Statement has been changed to 1, the default value of the group_replication_ssl_mode variable has been changed to REQUIRED, and the default value of group_replication_recovery_use_ssl variable has been changed to ON.
  For more information, see Replication Security. (WL #16438)
• To enable enhanced replication capabilities by default, the default values of gtid_mode has been changed to ON. Additionally, the default value of enforce_gtid_consistency has been changed from ON.
  For more information, see Replication with Global Transaction Identifiers. (WL #8602)

Spatial Data Support

• Geometry functions now support the inclusion of CRS URN in implicit JSON representation, ensuring that spatial reference system information is preserved when converting geometries to GeoJSON. This enhancement enables accurate representation and updating of geometry values in JSON duality views. It also affects the ST_AsGeoJSON function with options 2 and 4, which now always includes a CRS URN. With this update, GeoJSON resulting from an implicit conversion from a geometry type will have an extra "crs" attribute, providing a more comprehensive and accurate geometric representation.
  See Spatial GeoJSON Functions for more information. (WL #16929)

sys Schema Notes

• Errors encountered by the sys.ps_trace_thread stored procedure were not handled properly.
  Our thanks to Artem Avetian for the contribution. (Bug #34537707)
• The sys schema stored procedure, revoke_schema_privileges_from_all_accounts_except() is now installed by default. This procedure enables you to revoke privileges on a schema, for all users, except for a defined list of exceptions.
  See The revoke_schema_privileges_from_all_accounts_except() Procedure for more information. (WL #16996)

Thread Pool Notes

• MySQL Thread Pool now supports automatic hardware-based configuration, allowing it to detect the underlying server hardware and set optimal defaults for better performance. The plugin automatically adjusts its settings based on the number of VCPUs available on the system, and users can still make changes when needed. Additionally, the plugin automatically validates and correct misconfigured parameters, and logs warnings when values are auto-corrected.
  The following system variables are affected:
  • thread_pool_size
  • thread_pool_max_transactions_limit
  • thread_pool_query_threads_per_group
  • thread_pool_algorithm
  • thread_pool_max_unused_threads
    (WL #16621)

Bugs Fixed

• Logical clock handling for CREATE ... SELECT statements has been improved to ensure accurate logical clock values in the replica relay log. (Bug #38383106)
• Several updates were made to the generated API documentation.
  Our thanks to Daniël van Eeden for the contribution. (Bug #38341735)
• The server sometimes processed table references incorrectly. (Bug #38001000)

Original source

Note

These release notes were created with the assistance of MySQL HeatWave GenAI.

Authentication Notes

• The LDAP SASL authentication plugin could not be installed successfully when the Option Tracker component was already installed. (Bug #37983282)

Character Set Support

• Important Change: IFNULL() used in a LIKE clause raised ER_CANT_AGGREGATE_2COLLATIONS. This was due to the fact that IFNULL() returns a binary collation and the collation derivation NONE, but NONE was not allowed in this context, leading to the error.
  We solve this problem by making a slight change in semantics such that we lower the strength of collation derivation NONE to be less than the strength of any other collation derivation. This means that, when an expression with derivation NONE is used, the other operand determines the comparison collation to be used by LIKE. This should have minimal impact on existing functionality, and has required no changes in our existing test suite.
  We also change collation aggregation to consider only collations with the same (and highest) strength when determining the result collation. We also rename the previous IGNORABLE derivation to NULL, since it is used only for nulls, and since it is no longer ignorable given that NONE now has lesser strength.
  In addition, we leave a gap in the strength values for the former NONE strength, in order that as few COERCIBILITY() calls as possible return values that are incompatible with the previous implementation.
  See Collation Coercibility in Expressions, for more information. (Bug #37285902)
• CREATE TABLE with a generated column expression, such as a CHECK constraint, which referenced a non-ASCII identifier caused a syntax error if the current client character set was not compatible with UTF-8 (for example, GBK). (Bug #30453221)

Compilation Notes

• Group Replication: Defining TASK_EVENT_TRACE in gcs_profile.in broke the build of GCS/XCOM in MySQL Server's Group Replication GCS component, returning errors similar to member reference type 'connection_descriptor *' is a pointer; did you mean to use '->'? (Bug #38042851)
• macOS: It is now possible to compile the server on MacOS using -DWITH_KERBEROS.
• macOS: Aligned the buffer used for reading status variables. This fixes a potential issue with MacOS/M1 platforms.
• Upgraded the bundled libcurl library to version 8.14.1. (Bug #38042758)
• MySQL Server now supports CMake 4, ensuring compatibility with future CMake versions where support for versions prior to 3.10 is expected to be discontinued. (Bug #38027636)
• Use -DWITH_PROTOBUF=system for relevant .proto files, which can reduce build times. (Bug #38022751)
• Removed a warning in gcs_xcom_networking.cc. (Bug #38021787)
• The system bzip2 library is now located by cmake before linking with -lbz2. (Bug #38005363)
• MySQL Server now requires GCC version 11 or later to compile.
  As part of this change, support for ARM systems using RHEL7 is removed in this release, due to the lack of availability of GCC 11 on that platform. (Bug #38004285)
• Removed workaround for old versions of CMake.

Note

The minimum version of CMake supported for building MySQL is 3.14.6.

(Bug #37901122)

• The included ICU library has been upgraded to version 77-1. (Bug #37870791)
• The included zstd library has been upgraded to version 1.5.7. (Bug #37869972)
• Warnings raised by Clang 20 for non-trivially copyable types, deprecated literal operators, and incorrect main function declarations are no longer generated. (Bug #37785251)
• Removed warnings observed when compiling the server with Clang 20. (Bug #37785251)
• Disabled clang::musttail when building with GCC 15. (Bug #37776018)
• Worked around an issue with list handling in certain older versions of CMake. (Bug #37709169)
• Some compiler features tests did not pass when building with -fprofile-use. (Bug #37707556)
• On RHEL 9, MySQL Server is now built using GCC 14. (Bug #37702396)
• The version of Boost needed to build MySQL has been raised from 1.85.0 to 1.87.0. (Bug #37403602)
• The linker tried to use the empty Cmake variable ${ICU_LIBRARIES}, even though the correct library (ext::icu) was already linked elsewhere. (Bug #36524167)
• Compiling on macOS looked explicitly for [email protected] but now looks for the generic openssl symlink instead, to allow for openssl@3. (Bug #35468370)
• On MacOS, silenced deprecation warnings generated by Xcode 14; this includes suggestions to use snprintf(3) instead of sprintf(3), and warnings about possible loss of precision when converting from 64-bit to 32-bit integers. (Bug #34776172)

Component Notes

• When a UDF registered by a component is running, it cannot be unregistered, and so the component cannot be uninstalled (UNINSTALL COMPONENT is rejected with an error). When the component was uninstalled twice while the UDF was running, the first attempt failed (as expected), but the second one succeeded, causing the library to be unloaded, leading to issues with the UDF and possibly an unplanned server exit.
  This occurred because the component's deinitialization function cleared the container of registered UDFs even when it did not succeed, so the second uninstallation skipped deregistration. We fix this by storing the status of previous registrations in such cases. (Bug #35772996)
• The MySQL server's dynamic loader has been improved such that, when loading a component, its services are no longer registered until after the component's init() method has been called. This has the following implications for component behavior:
  • For developers, it is no longer necessary to perform sanity checks to ensure that needed data structures are present before calling the component's service methods.
  • For users, components should be more reliable especially in situations where components are loaded and unloaded with great frequency.
    See MySQL Components, for more information. (WL #16918)

Configuration Notes

• Microsoft Windows: An in-place upgrade of MySQL Server using MySQL Configurator failed when a Windows service name other than the default had been used. (Bug #37917039)
• Microsoft Windows: When upgrading a server from 8.0 to a higher series, MySQL Configurator did not persist custom server settings in the my.ini file. (Bug #37481548)
• Microsoft Windows: When upgrading a MySQL Server using MySQL Configurator, the process hung in the "Starting the server and upgrading system tables" step if a custom error log name was used. (Bug #37463478)
• Microsoft Windows: MySQL Configurator allowed additions of duplicate users, such as users defined with the same user name and host name, and then created them as different users on the server. With this fix, in GUI mode, a duplicate user is rejected with an error message, and in CLI mode, a duplicate user specified with --add-user is ignored. (Bug #37460190)
• Microsoft Windows: When used in CLI mode, MySQL Configurator now accepts a file path for the environment variable MYSQL_PWD, allowing the password to be specified in a file. (Bug #37460173)
• Microsoft Windows: When used in CLI mode, MySQL Configurator always rejected a root password shorter than 4 characters in length. With this fix, the restriction applies only for new configurations. In addition, the password option is no longer mandatory except for new configurations. (Bug #37460061)
• Microsoft Windows: As MySQL Enterprise Edition 9.4.0 includes a new The MySQL Enterprise Firewall Component, intended to be a replacement for the now deprecated MySQL Enterprise Firewall plugin, the MySQL Configurator now supports enabling the firewall component and upgrading to it from the firewall plugin:
  • For new server configurations: Users can choose to enable the new firewall component.
  • For server reconfigurations: Users can choose to disable the firewall, enable the new firewall component (if the firewall was not enabled before), keep the old firewall plugin, or upgrade the firewall plugin to the firewall component.
  • For server upgrades: If the firewall was enabled on the server before, users can choose to keep the old firewall plugin, or upgrade it to the firewall component.
    See MySQL Server Configuration with MySQL Configurator and MySQL Server Configuration with MySQL Configurator for details. (WL #16760)
• The default value of back_log is increased to 10000.
  See back_log for information on Linux configuration parameters which must be set to make MySQL Server resistant to connection bursts. (WL #16888)
• It is now possible to limit the maximum amount of physical memory used by MySQL Server when determining the default values of configuration variables, using server_memory. (WL #16938)

Deprecation and Removal Notes

• As part of ongoing work to transition from MySQL plugins to MySQL components, the API used to write MySQL plugins is now deprecated and subject to removal in a future version of MySQL. This change has the following effects:
  • The MySQL server's --early-plugin-load option is deprecated. Starting the server with this option now raises a deprecation warning.
  • A deprecation warning is now issued whenever a keyring plugin is loaded.
    Keyring Components Versus Keyring Plugins, provides a summary of the differences between keyring plugins and keyring components. See also: Key Migration Using a Migration Server, for information about migrating from a keyring plugin to a keyring component. (WL #16574)
• The system variable temptable_use_mmap, deprecated in version 8.0.26, is removed in this version. (WL #16745)
• The SQL functions MD5() and SHA1() are deprecated as of this release and scheduled for removal in a future release. (WL #16955)

Firewall Notes

• Packaging: This release includes a MySQL Enterprise Firewall component (see The MySQL Enterprise Firewall Component) intended to replace the firewall plugin, which is now deprecated and thus subject to removal in a future version of MySQL. The component implements most of the plugin's functionality, with the exception of account profiles, which are deprecated in the plugin and not supported by the component.
  If you are not already using MySQL Enterprise Firewall and wish to perform a clean installation of the firewall component, see Installing the MySQL Enterprise Firewall Component. For help with upgrading a current installation of the firewall plugin, see Upgrading to the MySQL Enterprise Firewall Component. Prior to upgrading the firewall plugin to the firewall component, you must convert any account profiles you may be using with the plugin to group profiles; Migrating Account Profiles to Group Profiles; provides assistance with this task.
  MySQL Enterprise Firewall is a commercial feature available as part of MySQL Enterprise Edition; see MySQL Enterprise Firewall, for more information. (WL #16570)

InnoDB Notes

• Memory management has been improved to prevent potential memory leaks, which could occur in certain error handling scenarios. (Bug #37826893)
• The information_schema.innodb_cmp_per_index table sometimes returned unknown values for database name and table name under certain conditions, such as when tables and indexes were evicted from the cache. (Bug #37820227)
• Fixed an issue relating to importing tables. (Bug #37621360)
• The default value of innodb_change_buffer_max_size has been changed to 5. This update aims to strike a balance between the benefits of change buffering for IO-bound workloads and the potential negative impact on in-memory workloads when a larger portion of the buffer pool is dedicated to change buffering. (WL #16968)

Installation Notes

• macOS: When mysqld was started without --plugin-dir, but with --basedir, where the base directory did not end in a slash character (/), and mysqld was configured to load the keyring_file component, server startup failed with errors. This caused problems for the macOS installer for MySQL, leading it not to set the password for the root account. (Bug #36816216)

References: See also: Bug #36398484.

• Debian packages for installing MySQL can now be run by users other than root. This helps prevent issues for Debian or Ubuntu systems that require rootless installations. (Bug #37765153)
• RPM and Yum repository installation are now supported for Red Hat Enterprise Linux and Oracle Linux 10. (Bug #37592019)

JavaScript Programs

• Under heavy memory usage, when attempting to execute a JavaScript stored procedure, the first context.parse() call in SP::init() raised an out-of-memory error. Subsequently, a second context.parse() call was made asynchronously; this could succeed if other sessions had cleared the memory in the meantime. This meant that an error was reported by the MLE component's error handler, but—since the second call succeeded—MLE returned a status code indicating success. Now in such cases, CALL is rejected with an out-of-memory error. (Bug #37952656)
• Upgraded the bundled version of Graal Maven to 24.2.1.0.1. (Bug #37938310)
• The included GraalVM and the Truffle library used by the MLE component have been upgraded to versions 23.1.7 and 24.2.1, respectively. (Bug #37833200)
• When mysqldump was from a MySQL 9.2.0 or later distribution, mysqldump --routines did not work properly with servers from previous versions, because it could not find the Information Schema LIBRARIES table. Now in such cases, mysqldump skips this table and does not try to dump it. (Bug #37498680)
• The MySQL BIT type is now supported for use in JavaScript stored routines. For type conversion rules and other information, see JavaScript Stored Program Data Types and Argument Handling. (WL #16885)
• JavaScript programs using the MySQL Multilingual Engine component now supports libraries written in WebAssembly (https://webassembly.org/). Such a library can be created using CREATE LIBRARY ... LANGUAGE WASM and the hexadecimal or base64-encoded representation of a compiled WebAssembly program.
• WebAssembly libraries are supported as a feature by the MySQL Option Tracker component (MySQL Enterprise Edition).
• JavaScript language support provided by the MySQL Multilingual Engine component now conforms to the ECMAScript 2024 Language Specification (ECMA-262, 15th Edition) as shown at https://262.ecma-international.org/15.0/. (WL #16887)

JSON Duality Views

• MySQL now supports JSON duality views, which provide a way to expose data stored in relational tables as JSON documents. Such views can be created, altered, dropped, and viewed using CREATE JSON DUALITY VIEW and ALTER JSON DUALITY VIEW (both implemented in this release); DROP VIEW and SHOW CREATE VIEW now work with JSON duality views as well as SQL views.
• For users of MySQL Enterprise Edition, JSON duality views can be modified using INSERT, UPDATE, or DELETE statements, so that updates of such a view cause corresponding updates on the base table. DML operations on JSON duality views are supported in MySQL Enterprise Edition only.
• The SELECT from a JSON duality view includes a 128-bit hash value generated for the document as part of an ETAG() field in the metadata section.
• Information about existing JSON duality views can be obtained from Information Schema tables: JSON_DUALITY_VIEWS, JSON_DUALITY_VIEW_TABLES, JSON_DUALITY_VIEW_COLUMNS, JSON_DUALITY_VIEW_LINKS.
• For users of MySQL Enterprise Edition, the JSON duality views feature is supported by the Option Tracker component.

Logging Notes

• MySQL Server now logs the total number of logical CPUs and physical memory accessible to the server in the error log. This information is always logged, regardless of the log verbosity. (WL #16940)

Packaging Notes

• On Fedora 24, Oracle MySQL RPMs could not be installed due to package conflicts when MariaDB was already installed on the system. With this fix, the conflicts are not properly handled to allow MySQL Server to be installed successfully. (Bug #37798784)

Performance Schema Notes

• The internal pfs_get_thread_statement_locker_vc() function contained debugging code that was inadvertently included in release builds. (Bug #37743667)

Vector Data Type

• The VECTOR_TO_STRING() function did not set its output collation correctly. (Bug #37815490)

Functionality Added or Changed

• Important Change: Added the mysql client --commands option, which enables or disables most mysql client commands.
  This option is disabled by default. To enable it, start the client with --commands or --commands=ON.
  For a complete list of all commands affected by this option, and additional information, see mysql Client Options. (WL #16949)
• InnoDB: To improve debugging, the buf_page_t and buf_block_t structure's metadata is now printed to the error log. (Bug #35115629)
• Group Replication: Added the error ER_GRP_RPL_APPLIER_THD_KILLED, to distinguish when the applier thread has been terminated using SQL KILL, rather than stopped by an error. (Bug #37764717)
• Increased the historical 1024-byte limit when printing the current query during signal handling to 1073741824 (1024 * 1024 * 1024). (Bug #37603354)
• Binary packages that include curl rather than linking to the system curl library have been upgraded to use curl 8.14.1. (Bug #37389565)

Bugs Fixed

• Important Note; Group Replication: The Group Communication System (GCS) handles Group Replication communication between members, and keeps track of the group membership and connections between all group members. Membership tracking includes the membership's current and previous two iterations. When a member leaves, the remaining members in the group keep a record of the departing member but stop communicating with it until it returns. For example: The group contains members M1, M2, and M3. M3 leaves the group; M1 and M2 stop communicating with M3. When a new member (M4) joins, it learns the previous iterations of this group's membership and attempts to communicate with all servers, including those from previous iterations (in this case, M3).
  In the event that some previous servers were gone and did not return, the new member continuously tried to connect to the missing servers; in some conditions, these ongoing connection attempts could introduce network latency affecting group member communication.
  In order to avoid this issue, connections to servers that belong to iterations of the group membership are now stopped after 5 minutes, which should be sufficient time to re-establish valid connections without imposing a continuing impact on group communication. (Bug #37704514)
• Performance: Redundant conditions in some queries optimized away in MySQL 8.0 were no longer removed in later versions, leading to a significant drop in the performance of such queries. (Bug #117907, Bug #37808260)

References: This issue is a regression of: Bug #30112096.

• NDB Cluster: Following an upgrade from NDB 8.0 to NDB 8.4, all data nodes in the cluster underwent an unexpected simultaneous restart. This occurred when the transaction coordinator had no scan state, leading to protocol timeout; the resulting misalignment in protocol states caused data nodes to shut down unexpectedly. This is fixed by extending existing handling of an unexpected SCAN_NEXTREQ signal to cover the case when the scan is already stateless. (Bug #37994985)

References: This issue is a regression of: Bug #37022901.

• NDB Cluster: Concurrent ALTER TABLE statements could cause delays of up to 100 * TransactionDeadlockDetectionTimeout before failing with a lock wait timeout when a client held a conflicting row lock, stalling the accompanying get_commit_count() call. The retry mechanism has been adjusted to identify locking issues sooner. (Bug #37955025)
• InnoDB: A check whether the table is referenced by a foreign key was performed for every row update, even when it was not required.
  The check is no longer performed unless it is required. (Bug #37867653)
• InnoDB: The return value from pthread_setaffinity_np was not properly interpreted when setting processor affinity for threads during buffer pool creation. As a result, no error was returned even if setting affinity failed. (Bug #37825544)
• InnoDB: When rebuilding a primary key, the server sometimes encountered issues when duplicates were present, potentially leading to the server stopping unexpectedly.
  Our thanks to Xizhe Zhang and the team at Alibaba for the contribution. (Bug #37822992)
• InnoDB: Fixed an issue relating to dropping columns that were part of an index. (Bug #37726881)
• InnoDB: MySQL Community Edition binaries included unnecessary OpenTelemetry symbols, due to unprotected static arrays. (Bug #37689163)
• InnoDB: The double write buffer was unnecessarily large. When calculating the number of segments per file, one extra segment was always added, whereas an extra segment should be added only if the number of Double_write instances is not divisible by the number of dblwr files. (Bug #37684656)
• InnoDB: Fixed an issue relating to DELETE operations. (Bug #37478594)
• InnoDB: Creating a secondary index on a VARCHAR column could allocate more memory than configured, with the amount allocated being directly related to the value of innodb_ddl_buffer_size, leading to errors similar to ERROR 1136 (21S01): Column count doesn't match value count at row 1. (Bug #37233273)
• InnoDB: Fixed an issue relating to indexing spatial datatype columns. (Bug #36682518)
• InnoDB: The temptable handler did not terminate cleanly during server shutdown, potentially causing the server to terminate unexpectedly. This issue has been addressed by implementing a mechanism to track and properly clean up temptable objects associated with each thread. (Bug #36538081)
• InnoDB: A long semaphore wait crash could occur when a redo log consumer lagged behind after a failed MySQL Enterprise Backup incremental backup, preventing the redo log writer thread from advancing. Error messages were returned similar to the following:
  [Warning]
  [MY-013934]
  [InnoDB]
  Redo log writer is waiting for MEB redo log consumer which is currently reading LSN = 23335640211468 preventing reclamation of subsequent portion of the redo log.
  Consider increasing innodb_redo_log_capacity.
  (Bug #36330455)
• InnoDB: Fixed an issue relating to range queries on tables. (Bug #31360522)

References: See also: Bug #38063122.

• Partitioning: Truncating a partition was rejected with a duplicate entry error when the partition ID exceeded INT_MAX, preventing creation of new partitioned tables. To mitigate this issue, the Table_partition_values_pk class constructor now uses ulonglong instead of int for the object ID. (Bug #35912852)
• Replication: When using replication in a chain, CREATE TABLE ... AS SELECT sometimes resulted in inconsistent entries in the binary log, potentially causing replication to break on downstream servers. Error messages related to this issue included errors resulting from missing START TRANSACTION statements in the log. (Bug #37986380)
• Replication: Fixed an internal memory management issue in libs/mysql/binlog/event/event_reader.cpp. (Bug #37371443)
• Replication: During semisync replication, when the length of the binary log suffix exceeded six digits (.999999), so that the next log file became—for example—mysql-bin.1000000, the replication protocol unexpectedly changed from semisynchronous to asynchronous.
  Our thanks to Wuhao Cao and Karry Zhang and the team at Alibaba for the contribution. (Bug #115861, Bug #113813, Bug #37024069, Bug #36246779)
• Group Replication: The Gr_empty_consensus_proposals_count system status variable was not updated as expected. (Bug #37937927)
• Group Replication: In an unstable network environment, a Group Replication InnoDB Cluster with group_replication_paxos_single_leader=ON experienced several long-running transactions that became stuck in the waiting for handler commit state. As a consequence, group_replication_set_as_primary() was forced to wait, which in turn blocked other incoming queries and rendered the cluster unwritable.
  The issue involving long-running transactions stuck in waiting for handler commit occurred as follows: During an intermittent network partition, a secondary node incorrectly assumed the leader role due to outdated or inaccurate membership information. This resulted in conflicts in synode number allocation, causing transactions originating from the primary node to remain incomplete.
  We fix this by making sure that a secondary node always reflect the latest, accurate state before pushing the view message to Paxos. This ensures that outdated or inconsistent membership information does not lead to conflicts in leadership or synode number allocation. (Bug #37764970)

References: See also: Bug #117424, Bug #37237959, Bug #37645674.

• NDB Cluster APIs: Excluding a VARCHAR column from an event definition resulted in an Invalid schema object version error. (Bug #37766391)

References: See also: Bug #31848270.

• RPM installations on Fedora 24 could not be completed because conflicting packages were pulled from upstream. This fix adds the proper obsoletes to block the conflicting packages. (Bug #37976913)
• In certain cases, expr BETWEEN expr AND expr (where expr is a complex expression) led to an assert in sql\sql_base.cc. (Bug #37952274)

References: This issue is a regression of: Bug #113506, Bug #36137690.

• Some triggers which called stored routines did not always execute correctly. (Bug #37915445)
• The internal function transform_table_subquery_to_join_with_derived() did not restore the current query block properly following invocation, leading to an assert. (Bug #37884336)

References: See also: Bug #37832605.

• Removed a potential memory leak in item_cmpfunc.cc. (Bug #37883669)
• Using the HLL() function (MySQL MySQL HeatWave only) with a CAST() operation to UNSIGNED on a column, along with a STRCMP() operation and a NULL in the WHERE clause, now behaves as expected. Previously, this returned errors similar to Assertion `!null_value || is_nullable()' failed. (Bug #37839325)
• A view using a CTE which contained a subquery was not always handled correctly. (Bug #37832605)
• Removed a performance regression introduced by work done in MySQL 9.2.0 on the caching SHA-2 authentication plugin. (Bug #37523469)
• UPDATE ... SET ... could not always be rolled back successfully. (Bug #37489167)
• Multiple DEFAULT column expressions in CREATE TABLE statements were not always handled correctly. (Bug #37436220)
• Queries having a correlated subquery which performed aggregation were sometimes (incorrectly) rejected with duplicate key errors during execution.
  This issue was introduced by a previous fix which removed the restoration of the original reference slice (set_ref_item_slice(REF_SLICE_SAVED_BASE)) during JOIN::cleanup() execution under the assumption that this was not necessary. As a result, temporary table field references from previous executions were not cleared, leading to attempts to insert duplicate keys into temporary tables triggering the error Can't write; duplicate key in table.
  We fix this by restoring the original reference slice during cleanup, ensuring that any stale references are discarded. (Bug #37415167)

References: See also: Bug #32141711. This issue is a regression of: Bug #35856247.

• On some Windows systems, after installing MySQL Server 9.1 with the MSI installer, the server failed to start. This was because those systems did not have Visual C++ Redistributable v.14.42 or later installed. The MySQL Server MSI installer now checks for that requirement and refuses to start installation unless it is satisfied. (Bug #37365476)
• Unquoted semicolon characters (;) within comments were not always flagged as errors, in spite of the fact that they are not allowed. (Bug #37117875)

References: See also: Bug #38063286.

• Removed a potential memory leak from the keyring_aws plugin. (Bug #36684413)
• A query using a nested aggregate function which contained a subquery was not always properly handled. (Bug #36421727)
• When attempting to transform a subquery to a derived table, certain cases were not always considered. (Bug #36421710)

References: This issue is a regression of: Bug #36921175.

• Setting max_join_size led to improper processing of some nested queries. (Bug #35625769)
• The transform of a scalar subquery into a join with a derived table where the subquery is in the SELECT list and the containing query is implicitly grouped should be allowed, but was rejected when the subquery_to_derived optimizer switch was enabled. (Bug #35150438)
• EXPLAIN ANALYZE FORMAT=JSON did not handle queries with subqueries correctly when explain_json_format_version was equal to 1.
  Our thanks to Peiyuan Liu and the Tencent team for the contribution. (Bug #117995, Bug #37285902)
• It was possible to use PERSIST or PERSIST_ONLY with SET TRANSACTION ISOLATION LEVEL, even though this should not be allowed, and later caused errors on server restart. Now attempting to do so causes the statement to be rejected with ER_GRP_RPL_UNSUPPORTED_TRANS_ISOLATION. (Bug #115619, Bug #36854635)
• The query rewrite plugin (see The Rewriter Query Rewrite Plugin) did not work properly when the server was run with autocommit=OFF. (Bug #115437, Bug #36784795)
• MIN(), when used as a window function, did not ignore nulls as expected. (Bug #113631, Bug #36182490)
• It was possible in a window frame specification using RANGE frame units to have an ORDER BY expression containing RAND(). This broke an invariant in the RANGE frame specification; that the values be monotonically ascending or descending. We solve this issue by disallowing a non-deterministic ORDER BY expression when combined with RANGE units; this is now rejected with an error.
  This also fixes a related issue in which RANGE unit comparison failed when computing the frame limits for BETWEEN CURRENT ROW AND after_value FOLLOWING where the row value being compared was unsigned and could be less than after_value, leading to underflow and a possible incorrect result. In such cases, we now reject the operation with an error.
  See Window Function Frame Specification, for more information. (Bug #111510, Bug #35521787)

Original source

Account Management Notes

• It was possible in some cases to grant a user the EXECUTE privilege, but subsequently to be unable to revoke it from the same user. (Bug #37570206)

Audit Log Notes

• <COMMAND_CLASS> was not populated for <NAME>Execute</NAME>. For more information, see Logging Specific Event Classes. (Bug #36686351)

Authentication Notes

• This release introduces a MySQL component, component_keyring_hashicorp, which takes the place of the HashiCorp Vault keyring plugin; that plugin and its associated system variables are now deprecated (and subject to removal in a future version of MySQL). Like the plugin, the component allows communication with HashiCorp Vault for backend storage, and supports HashiCorp Vault AppRole authentication, but has been reimplemented using the MySQL component architecture.
• For more information, including information about installation and configuration of the HashiCorp Vault keyring component, see Using the HashiCorp Vault Keyring Component. For additional information about MySQL keyring components, see Keyring Component Installation, as well as Security Components and Plugins. For general information about MySQL components, see MySQL Components.
• component_keyring_hashicorp plugin is an extension included in MySQL Enterprise Edition, a commercial product. To learn more about commercial products, see https://www.mysql.com/products/. (WL #16505)

Compilation Notes

• Group Replication: The OpenSSL Engine interface is deprecated, and is no longer being included in OpenSSL v3 main packages by some Linux distributions, including Fedora.
• To avoid build issues, the usage of the OpenSSL Engine interface by the Group Communication System (GCS) is now restricted to OpenSSL versions previous to 1.1. (Bug #37475769)
• Linux: Use /usr/bin/gcc (GCC 14.2.1) when building the server on Oracle Linux 10. (Bug #37616148)
• Linux: Use --experimental_allow_proto3_optional when building with protoc version 3.14 or earlier. (Bug #37579947)
• Microsoft Windows: The server could not be built using Visual Studio 17.13.1 with MSVC_CPPCHECK enabled. (Bug #36925076)
• Upgraded the bundled Curl library to version 8.12.1. (Bug #37633587)
• Disabled the -ftls-model=initial-exec option when compiling MySQL on FreeBSD. (Bug #37613105)
• Abseil could not be built on FreeBSD. (Bug #37611924)
• Do not read share/charsets/Index.xml when running comp_err. (Bug #37569683)
• The file mysql_version.cmake was included multiple times. (Bug #37559512)
• Removed the unused file strings/utr11-dump.cc. (Bug #37549844)
• The bundled version of opentelemetry-cpp was upgraded to version 1.19.0. (Bug #37506554)
• Fixed a large number of warnings generated by clang-tidy. (Bug #37471922)
• include/my_systime.h included std::chrono, which was unneeded, and which has now been removed. (Bug #37458343)
• In order to use xxhash functions independently from the lz4 library (bundled or source), we compiled xxhash.c into our own binaries, which required using a great many CMake directives. Instead, we now build an interface library for xxhash, and link with that wherever such functions are used. (Bug #37417386)
• Use xxHash-0.8.2 from GitHub rather than the version bundled with lz4. (Bug #37387318)
• The bundled version of opentelemetry-cpp was upgraded to version 1.18.0. (Bug #36708755)
• Added documentation for unused bytes at the end of Protocol::ColumnDefinition41.
• Our thanks to Daniyaal Khan for the contribution. (Bug #117346, Bug #37541403)

Component Notes

• Important Change: As of this release, MySQL Enterprise Data Masking and De-Identification, part of MySQL Enterprise Edition, is now known as MySQL Enterprise Data Masking.
• Our documentation, beginning with the 9.2 edition, has been updated to reflect this change.
• For more information, see MySQL Enterprise Data Masking. (WL #16721)
• Group Replication: Some of the entries in package specification files for normal and debug files for the Flow Control Statistics component were misplaced.
• The Flow Control Statistics component is available as part of MySQL Enterprise Edition. For more information, see Group Replication Flow Control Statistics Component. (Bug #37486491)
• Group Replication: Added the Group Replication Primary Election component, which makes it possible to specify the most-up-to-date selection method for choosing a new primary in the event of failover. The Group Replication plugin is a prerequisite for this component, which must be installed on each group member. In addition, the group_replication_elect_prefers_most_updated.enabled system variable must be set to ON, on each group member, in order for the component to function.
• This component also provides two status variables for monitoring purposes. Gr_latest_primary_election_by_most_uptodate_members_trx_delta was the difference in the number of transactions between the new primary and the secondary most up to date with it, when the most-up-to-date primary selection method was last used. Gr_latest_primary_election_by_most_uptodate_member_timestamp provides a timestamp for the most recent election of a new primary using the most-up-to-date selection method.
• When a new primary is elected, the component records the event in the log. This information in the log entry includes a timestamp, the UUID of the promoted secondary, and the method used to select the new primary: either the most-up-to-date method (and how many transactions this secondary was behind the primary), or weighting (with the weight that was assigned to the secondary chosen).
• For more information about this component, see Group Replication Primary Election Component.
• This component is part of MySQL Enterprise Edition, a commercial offering. See MySQL Enterprise Edition, for more information. (WL #16432)
• The following enhancements have been made in this release to the MySQL Option Tracker component, part of MySQL Enterprise Edition:
  • Each feature that supports the Option Tracker now provides a global status variable named option_tracker_usage:feature_name, which provides a count of the number of times a feature has been used. This variable is provided whether or not the Option Tracker component is installed.
  • You can view these status variables using SHOW GLOBAL STATUS LIKE 'option_tracker_usage%' or by selecting from the Performance Schema global_status table.
  • The Boolean used key in the JSON format employed for usage data has been replaced with a counter, usedCounter. After upgrading to this release, no used members are added or updated in this data by the Option Tracker.
• See Option Tracker Status Variables, as well as Option Tracker Supported Components, for more information. (WL #16721)

Configuration Notes

• Microsoft Windows: The --install-sample-database option was ignored by the MySQL Configurator CLI. (Bug #37701034)
• Microsoft Windows: MySQL Configurator CLI did not execute any actions specified with the --action option, except for configure. (Bug #37473745)
• Microsoft Windows: During an upgrade from MySQL 9.1.0 to 9.2.0, the MySQL Configurator failed to find the my.ini configuration file and required manual selection of its file path. (Bug #37468826)
• Microsoft Windows: Upgrading an existing server with a non-default instance port by the MySQL Configurator CLI failed, unless the --old-instance-protocol option was used in the command. (Bug #37459238)

Deprecation and Removal Notes

• NDB Cluster: The --restore-privilege-tables option for ndb_restore, which was deprecated in NDB 8.0.16, has now been removed. (Bug #36298807)
• Replication: The replica_parallel_workers server system variable can no longer be set to 0; the minimum permitted value is now 1. (WL #13957)
• The system variable innodb_undo_tablespaces, deprecated in MySQL 8.0.14, has been removed. (WL #16746)
• The system variables innodb_log_file_size and innodb_log_files_in_group, deprecated in MySQL 8.0.30, have been removed. (WL #16743)
• The Version Tokens plugin, which was deprecated in MySQL 9.2.0, has been removed in this release. (WL #16614)

Doxygen Notes

• Addressed issues in the server source code documentation as noted here:
  • The index page linked to the MySQL 8.0 Manual. This fix avoids the versioning issue here by using an unversioned link instead.
  • In protocol_classic.cc, the parameter type and flag descriptions were unclear, and have been improved.
• Our thanks to Daniël van Eeden for the contributions. (Bug #117391, Bug #117503, Bug #37559971, Bug #37607749)
• Addressed the following issues in the server source code documentation for COM_STMT_PREPARE response packets:
  • The payload specification showed the wrong condition for checking the warning count.
  • Removed extraneous pipe characters (|) from the first example.
• Our thanks to Kanno Satoshi for the contribution. (Bug #117373, Bug #37552681)
• Addressed the following issues in the server source code documentation:
  • Corrected misspelling in the MySQL Client/Server Communication protocol documentation for AuthSwitchRequest.
  • Table for distinguishing OK packets from EOF packets used the wrong operator; this has now been corrected.
  • Documented previously undocumented behavior regarding the sending of default values in COM_FIELD_LIST.
• Our thanks to Daniyaal Khan for the contributions. (Bug #117325, Bug #117374, Bug #117596, Bug #37534532, Bug #37552684, Bug #37645678)

SQL Function and Operator Notes

• Important Change: When an SQL function is improved from one release to the next, it may throw SQL errors in situations in which it previously did not. If this happens in a table's constraints, default expressions, partitioning expressions, or virtual columns, the table could not be opened. This prevented both analyzing the problem (using, for example, SHOW CREATE TABLE) and addressing it (such as with an ALTER TABLE ... DROP ... statement).
• Now, on server upgrade, we scan the data dictionary for tables that use any of the features just mentioned. We then try to open such tables, and if we fail to do so, we alert the user. This patch addresses this. The --check-table-functions server option introduced in this release helps to address this problem by making it possible to specify the server's behavior when encountering an error with such a function. Set this option to WARN in order to log a warning for each table which the server could not open; setting it to ABORT also logs these warnings as WARN, but aborts the server upgrade if any issues were found.
• ABORT is the default; this enables the user to fix the issue using the older version of the server before upgrading to the new one. WARN flags the issues, but allows the user to continue in interactive mode while addressing the problem. (Bug #36890891)
• References: See also: Bug #37009318. This issue is a regression of: Bug #98950, Bug #98951, Bug #31031886, Bug #31031888.

INFORMATION_SCHEMA Notes

• Fixed a performance issue in the PROCESSLIST table. (Bug #36778475)

InnoDB Notes

• InnoDB now uses simplified APIs for saving and restoring level one progress during parallel scans. (Bug #37517125)
• InnoDB now supports container-aware resource allocation, allowing it to adhere to the restrictions imposed by the container. The default values of InnoDB configurations are now calculated based on the logical CPUs and physical memory allocated by the container, rather than relying on system-wide resources.
• The values for the following system variables are calculated and set based on those resource limits:
  • The values of the following are calculated based on the number of logical CPUs:
    • innodb_buffer_pool_instances
    • innodb_page_cleaners
    • innodb_purge_threads
    • innodb_read_io_threads
    • innodb_parallel_read_threads
    • innodb_redo_log_capacity (value set only if --innodb-dedicated-server is enabled.)
    • innodb_log_writer_threads
  • The values of the following are calculated based on the available memory:
    • temptable_max_ram
    • innodb_buffer_pool_size (value set only if --innodb-dedicated-server is enabled.)
• See Container Detection and Configuration. (WL #16484)

JavaScript Programs

• Important Change: JavaScript stored programs now fully support the DECIMAL type, including its alias NUMERIC; can now be used with JavaScript programs as input arguments, output arguments, prepared statement bind() parameters, and return values.
• In order to maintain precision, MySQL DECIMAL is converted to JavaScript String by default, but it is possible to override this behavior, causing it to be converted to Number instead by setting the value of the decimalType option to NUMBER (or mysql.DecimalType.NUMBER).
• It is possible to convert JavaScript Boolean, Number, String, and BigInt values to DECIMAL (or NUMERIC). Trying to convert any other JavaScript type to a MySQL decimal type is not supported, and is rejected with an error.
• See Conversion to and from MySQL DECIMAL and NUMERIC, for additional information and examples.
• JavaScript stored program support requires the Multilingual Engine Component (MLE), available with MySQL Enterprise Edition. See Multilingual Engine Component (MLE). (WL #16747)
• Important Change: Dynamic import of JavaScript libraries is now supported using the await operator. This means you can use constructs such as that shown here to insure that libraries are loaded before using them:

let module = await import(`/db1/lib_${object_type}`)
return module.default.print()
// assume this method is defined for each lib

• See Using JavaScript Libraries, for more information and examples. (WL #16733)
• Upgraded the MLE component to use GraalVM Truffle version 24.2.0. (Bug #37668857)
• Importing a library with a global wait led to an internal error. (Bug #37425528)
• An issue with error handling led to an assert in sql/sql_class.cc. (Bug #36777428)
• The mle_session_reset() function has been enhanced with an optional string argument which takes one of the values "stderr", "stdout", or "output" to clear stderr, stdout, or both, respectively.
• When called without an argument, mle_session_reset() behaves exactly as it did in previous versions of MySQL: it clears both stderr and stdout, resetting the session time zone and clearing the stack trace as it does so; this removes any observable output from mle_session_state().
• For more information, see JavaScript Stored Programs—Session Information and Options.
• mle_session_reset() is provided by the MLE component, part of MySQL Enterprise Edition. See Multilingual Engine Component (MLE), for more information about this component. (WL #16660)
• MySQL JavaScript programs now provide localization and internationalization of numbers, dates, and other values by supporting the Intl global object. MySQL locales map to JavaScript locales by substituting a dash character for the underscore; for example, setting lc_time_names = "ja_JP" means that JavaScript returns "ja-JP" for the locale.
• It is also possible to override the session or default locale within a stored program by calling an object's toLocaleString() method or making use of one of the Intl formatting objects.
• The locale in effect the first time a JavaScript stored program is invoked in a given session remains the default locale for that stored program (unless it explicitly overrides the locale) until the session is reset, even after setting lc_time_names to a new value.
• For more information and examples, see JavaScript Localization and Internationalization. (WL #16709)
• This release includes a number of changes and additions relating to JavaScript library SQL:
  • The statements ALTER PROCEDURE and ALTER FUNCTION now accept a USING clause and so can add, replace, or remove a list of libraries imported by the named JavaScript stored procedure or stored function. See the descriptions of these statements for more information and examples.
  • CREATE LIBRARY now supports a COMMENT clause. This comment is shown in the output of SHOW CREATE LIBRARY and SHOW LIBRARY STATUS (see next item). It is also shown in the ROUTINE_COMMENT column of the Information Schema ROUTINES table; the LIBRARIES table also adds a LIBRARY_COMMENT column for displaying this value.
  • This release also implements a SHOW LIBRARY STATUS statement which provides basic information about one or more JavaScript libraries. Like SHOW PROCEDURE STATUS, this statement supports LIKE and WHERE clauses for filtering the output.
  • The ALTER LIBRARY statement implemented in this release makes it possible to update the comment for a JavaScript library.
  • The status variables Com_alter_library and Com_show_library_status, added in this release, provide counts of ALTER LIBRARY and SHOW LIBRARY STATUS statements, respectively.
• For general information about JavaScript stored programs, see JavaScript Stored Programs. See also Multilingual Engine Component (MLE). (WL #16737)

JSON Notes

• It was possible to create a table including a JSON column with DEFAULT '' under the default sql_mode, but the output from SHOW CREATE TABLE for this table, when run on another mysqld, resulted in the error BLOB, TEXT, GEOMETRY or JSON column ... can't have a default value, even when the sql_mode of the second mysqld was also non-strict.
• This issue did not occur with BLOB or TEXT columns. (Bug #116479, Bug #37219226)
• A query using WHERE EXISTS( SELECT ... FROM JSON_TABLE(...) ) did not return the expected result. (Bug #114897, Bug #3666073)

MySQL Enterprise Notes

• The Option Tracker component has added support for two MySQL features—the traditional MySQL Optimizer, and the MySQL Hypergraph Optimizer (available in MySQL HeatWave only).
• See Option Tracker Supported Components, for more information. (WL #16548)

Optimizer Notes

• Previous versions of MySQL, when the subquery_to_derived optimization was enabled, supported transformation into an inner or outer join with a derived table corresponding to a subquery using one of the quantified comparison operations =ANY (equivalent to IN) or <>ALL (equivalent to NOT IN) in the WHERE clause of a query. This release extends the functionality of this optimizer switch in two ways:
  • a. All such comparisons are now supported (>ANY, >=ANY, <ANY, <=ANY; >ALL, >=ALL, <ALL, <=ALL; =ANY, <>ALL).
  • b. The transformation of such comparisons is now supported in the SELECT clause and in the WHERE clause.
• For more information and examples, see Optimizing ANY and ALL Subqueries. (WL #13052)
• References: See also: Bug #37616992.

Performance Schema Notes

• The PERFORMANCE_SCHEMA service thread v7 was not exposed, preventing its use by components. (Bug #37579218)
• You can now configure a network namespace for your telemetry endpoints on Linux platforms. The following system variables are added:
  • telemetry.otel_exporter_otlp_traces_network_namespace
  • telemetry.otel_exporter_otlp_metrics_network_namespace
  • telemetry.otel_exporter_otlp_logs_network_namespace
• (WL #16735)

Functionality Added or Changed

• Important Change: Beginning with this release, it is no longer possible to downgrade between individual MySQL Innovation series releases, even within the same series. For example, were a version 9.3.1 to be released, it would not be possible after upgrading to it to downgrade back from MySQL 9.3.1 to 9.3.0. (Bug #37387488)
• Important Change: For platforms on which OpenSSL libraries are bundled, the linked OpenSSL library for MySQL Server has been updated to version 3.0.16. For more information, see OpenSSL 3.0 Series Release Notes and OpenSSL Security Advisory (11th February 2025). (Bug #36033684)
• Performance: The output process of the mysql client for binary values (printed as hexadecimal strings) in query results has been optimized, speeding up slightly the output of large result sets containing binary values. (Bug #37334107)
• Added Enterprise Linux 10 (EL10) support. (Bug #37592019)
• The signature of the handlerton drop_database_t API has been changed: instead of the database path, it now accepts the database name as a parameter. The change makes the API more efficient. (Bug #37191149)
• The mysqldump utility can now provide logical dumps of information about user accounts, writing the appropriate CREATE USER and GRANT SQL statements to the dump, when run with the --users command line option introduced in this release.
• You can cause the CREATE USER statements generated by mysqldump to be preceded by DROP USER by including the --add-drop-user option as well.
• It is also possible to include or exclude specific user accounts from the dump using either of the options --include-user or --exclude-user.
• For more information and examples, see the descriptions of the options cited, as well as mysqldump — A Database Backup Program. (Bug #28038954, WL #15658)
• The mysql client now displays query execution times with three decimal places of precision to show milliseconds.
• Our thanks to Marcelo Altmann for the contribution. (Bug #117270, Bug #37510263)

Bugs Fixed

• InnoDB: Under certain circumstances, Trx_by_id_with_min::insert() during get_better_lower_bound_for_already_active_id() could set an incorrect s_lower_bound value. (Bug #37548045)
• InnoDB: Fixed a potential memory leak in several places in the innobase code. (Bug #37403052)
• InnoDB: Under certain circumstances, MySQL could crash during shutdown due to pages which were still fixed or dirty. Errors similar to the following were logged: [ERROR] [MY-011908] [InnoDB] [FATAL] Page [page id : space = 46, page number = 75] still fixed or dirty [ERROR] [MY-013183] [InnoDB] Assertion failure: buf0buf.cc: 5889: ib::fatal triggered thread 139963705668608 (Bug #37391519)
• InnoDB: CHECK TABLE for spatial indexes did not verify the MBR against the geometry MBR stored in the clustered index record. This could result in incorrect behavior of spatial indexes. As of this release, CHECK TABLE EXTENDED verifies the MBR matches the MBR stored in the clustered index record. (Bug #37359538)
• InnoDB: Fixed an issue relating to pessimistic row update. Our thanks to Mengchu Shi and the team at Alibaba for the contribution. (Bug #37292404)
• InnoDB: The CHECK TABLE operation could incorrectly report corruption in spatial indexes. (Bug #37286473)
• InnoDB: Fixed an issue relating to InnoDB redo log recovery. (Bug #37061960)
• InnoDB: Fixed an issue relating to reading index_id values. (Bug #36993445, Bug #37709706)
• InnoDB: Under certain circumstances, an assertion failure occurred if the InnoDB engine performed unnecessary conversions for end range checks. This resulted in an error similar to the following: Assertion failure: lob0lob.cc: 897: trx == nullptr || trx->is_read_uncommitted() (Bug #35006212)
• InnoDB: Fixed an issue relating to lower_case_table_names. (Bug #32288105)
• InnoDB: Partition table indexes were not checked when retrieving a record count while that table's definition was being altered by another client session. The record count was executed without error. As of this release, the index is checked to ensure it is usable when retrieving a record count. (Bug #117459, Bug #37617773)
• InnoDB: Refactored code related to BPR_PCUR_* positioning for restore operations. (Bug #117259, Bug #37505746)
• References: This issue is a regression of: Bug #37318367.
• InnoDB: Changes made to innodb_spin_wait_delay in MySQL 8.0.30 negatively impacted performance. (Bug #116463, Bug #37212019)
• InnoDB: Under certain circumstances, using ALTER TABLE with INPLACE to modify the size of a column could result in an index which exceeds the valid size limit (767 bytes). This occurred for tables with a row format of Redundant or Compact and the row format was not explicitly defined in the table creation. As of this release, a validation is performed and an error returned by any ALTER TABLE, INPLACE operation which will result in an invalid index size. (Bug #116353, Bug #37168132)
• InnoDB: Fixed a memory leak in the Clone_persist_gtid thread. Our thanks to Baolin Huang and the team at Alibaba for the contribution. (Bug #107991, Bug #34454572)
• Partitioning: When inserting NOW() into a column not part of the partition key of a partitioned table, all partitions were retrieved, and no pruning occurred. (Bug #37397306)
• Replication: Removed a potential race condition in rpl_opt_tracker.cc. (Bug #37644518)
• Replication: When the log sanitizer analyzes relay log files, it first searches for a starting point (such as a rotation event or transaction end), but in some cases, it was possible for a binary log file containing a needed GTID to be deleted as having no relevant data; this resulted in the point-in-time recovery process hanging indefinitely while waiting for the missing GTID to be applied. Now in such cases, the analysis skips parsing of transaction boundaries until the start point is established. (Bug #37635908)
• Replication: In a source-replica setup, the replica encountered irregular failures of UPDATE and DELETE statements with ER_KEY_NOT_FOUND errors on the same table. (The replica's binary log and GTID records showed that the row required was committed, and had not been deleted or updated.) This occurred on the replica when the row-matching algorithm used was HASH_SCAN and two rows in the same table had the same CRC32 value.
• In the event of such a CRC32 collision, finding a matching CRC32 in the hash table does not guarantee that the correct row is being updated, so the algorithm iterates over any multiple entries having the same CRC32, and compares the full record for each of them in a loop; the problem occurred due to the fact that the logic to exit this loop was incorrect. This logic has now been corrected. (Bug #37462058, Bug #37731216)
• Replication: It was found during testing that it was possible to force the process responsible for termination of replica threads to access a deleted object. (Bug #37375269)
• Replication: The asynchronous_connection_failover_delete_source() function did not always perform as expected in all cases. (Bug #36479088)
• Replication: In some cases, the asynchronous_connection_failover_add_source() function did not perform as expected. (Bug #36479083)
• Replication: In some cases, MASTER_POS_WAIT() did not perform as expected. (Bug #36421684, Bug #37709187)
• Replication: The asynchronous_connection_failover_add_managed() function in some cases did not produce the expected result. (Bug #34648589)
• Replication: When the server was under a heavy write load, the binary log position for gtid_executed as shown in the Performance Schema log_status table did not match that of the gtid shown in the binary log file.
• We fix this by increasing the scope of the lock on the log_status table when querying it to ensure that transactions in the commit pipeline are completed. This ensures that a query against the log_status table waits until gtid_executed has been fully updated, thereby guaranteeing consistency with its position in the binary log. (Bug #102175, Bug #32442772)
• Group Replication: When a secondary joined the group, it might happen that all group members started to grow the value of the column COUNT_TRANSACTIONS_ROWS_VALIDATING column of the Performance Schema replication_group_member_stats table indefinitely. This impacted memory consumption in all group members, eventually leading to thrashing if not mitigated by restarting the secondary group member that triggered the behavior, or in some cases, by restarting the whole group.
• Analysis pointed to issues with the Group Replication start operation, which checks whether there are partial transactions on the group_replication_applier channel from previous group participation; if any are found, this channel is stopped after applying all complete transactions and its relay logs purged, and then the channel is restarted. After this, distributed recovery is performed, applying any missing data from group members.
• The issues arose when the Group Replication pipeline operation for stopping the group_replication_applier channel incorrectly stopped a periodic task from the certifier module, which caused some periodic internal operations not to take place. One of these tasks was the periodic sending of the committed transactions; this omission prevented garbage collection for certification, which in turn caused a continuous increase in COUNT_TRANSACTIONS_ROWS_VALIDATING in the Performance Schema replication_group_member_stats table.
• To solve this problem, we have taken steps to ensure that the pipeline operation for stopping the group_replication_applier channel no longer interferes with the certifier module, which also stops spurious values from being added for COUNT_TRANSACTIONS_ROWS_VALIDATING. (Bug #37613510)
• Group Replication: When running Group Replication, some transactions may not have write sets, as with empty transactions with GTID_NEXT specified, or DDL statements. For such transactions, Group Replication cannot check conflicts; thus, it is not known whether they can be applied in parallel, and for this reason, Group Replication follows a pessimistic approach, and runs them sequentially, potentially leading to an impact on performance.
• While DDL must be applied sequentially, there is no actual reason to force such behavior for empty transactions, so this fix makes it possible for empty transactions to be applied concurrently with other nondependent transactions. (Bug #37597512, Bug #37569333)
• Group Replication: Removed redundant GCS code which tested the same conditional variable twice in succession. (Bug #37538338)
• Group Replication: A group running group replication with a primary i1 and two secondaries i2 and i3 started to have intermittent issues because of high memory usage on the primary. The secondaries began reporting the primary as unreachable then reachable again, and the primary began reporting the secondaries as intermittently reachable then reachable as well. Following a period of such instability, the secondaries expelled the original primary (i1) and elected a new one (i2).
• Under these conditions, queries against the performance_schema.replication_group_members table on the former primary (i1) reported i1 as ONLINE and PRIMARY, i2 as ONLINE and SECONDARY, and i3 as ONLINE and SECONDARY for an extended period of time (12 hours or more) until the mysqld process was restarted on i1.
• The problems observed were found to have begun on the original primary (i1) when one of the secondaries was overloaded and began intermittently leaving and joining the group, its connections being dropped and recreated repeatedly on the primary server. During the reconnection process, the primary hung when trying to create the connection, thus blocking the single XCom thread. This was traced to the invocation of SSL_connect() on the XCom communication stack, which changed in MySQL 8.0.27 from asynchronous to synchronous form. When a node was overloaded, it might not respond to the SSL_connect() call, leaving the connecting end blocked indefinitely.
• To fix this, we now connect in a way that is non-blocking, and that returns in case of a timeout, leaving the retry attempts to the caller—in this specific case, the XCom thread when trying to reconnect to another node. (Bug #34348094, Bug #36047891)
• References: See also: Bug #37587252.
• The fprintf_string() function in mysqldump did not use the actual quote character for string escaping. (Bug #37607195)
• Cleaned up code in overflow_bitset.h. (Bug #37591520)
• Use std::string_view rather than std::string when looking up character sets and collations, which saves on memory allocation and deallocation. (Bug #37586193)
• Removed code left unused after a previous fix. (Bug #37574896)
• References: This issue is a regression of: Bug #28956360.
• Use std::string::starts_with() instead of the starts_with() function defined in strings/ctype.cc, and remove the latter function as no longer needed. (Bug #37568373)
• EXPLAIN did not always handle subqueries correctly. (Bug #37560280)
• Collation name aliases were sometimes handled in case-sensitive fashion. (Bug #37554688)
• References: This issue is a regression of: Bug #36878077.
• If a demangled function name exceeded 512 bytes in a stack trace, the function name was truncated and a newline was not printed.
• As of this release, long strings, such as filenames and demangled functions, are written directly to the output. (Bug #37543598)
• mysqldump did not escape certain special characters properly in its output. With this fix, mysqldump now follows the rules as described in String Literals. (Bug #37540722, Bug #37709163)
• Some operations on tables having functional indexes were not handled properly. (Bug #37523857)
• If a server was installed on an Enterprise Linux platform using RPM packages, after installing component_log_sink_json, trying to set log_error_services resulted in an error. This was due to a permission issue with the file path of the JSON log file, which has been fixed by this patch. (Bug #37508168)
• Attempting to install an unknown component using INSTALL COMPONENT was not always handled correctly. (Bug #37437317)
• For user input such as COLLATE utf8_bin we perform an alias lookup to find the actual collation (in this case, utf8mb3_bin). Now we use this name, rather than the input string, when reporting SQL errors. (Bug #37412963)
• Removed the internal binary_keyword variable, which was not actually used. (Bug #37408338)
• In libmysqld, errors were not correctly handled in udf_handler::add() for aggregates. (Bug #37398919)
• Removed the potential for undefined behavior in certain cases from the internal function check_if_server_ddse_readonly(). (Bug #37394933)
• The internal function recover_innodb_upon_upgrade() was no longer used, and has been removed. (Bug #37394850)
• The Audit Log plugin did not handle errors correctly when writing JSON output.
• See MySQL Enterprise Audit Plugin (Deprecated), for more information. (Bug #37370439)
• ER_SERVER_OFFLINE_MODE was not always handled correctly. (Bug #37355755)
• An UPDATE subsequent to an INSERT affecting a table which had a BEFORE INSERT trigger was sometimes rejected with a null value error when the INSERT had set a NOT NULL column to NULL, even though this should have been allowed by the server sql_mode in effect. (Bug #37337527)
• In some cases, components could not reuse the same connection for running multiple queries. (Bug #37286895)
• Improved error handling for stored routines. (Bug #37193011)
• Stored routines were not always invoked correctly in prepared statements. (Bug #37077424, Bug #37292797)
• Removed an error found in the preparation of stored functions. (Bug #36684438)
• Increased the size of SEL_ROOT::elements from uint16 to size_t. (Bug #36610878)
• Removed an issue with multibyte UTF8 handling. (Bug #36593253)
• An ORDER BY containing an aggregation was not always handled correctly. (Bug #36593244)
• An optimizer hint was ignored, unexpectedly requiring the use of FORCE INDEX, when querying a view that included a UNION. For more information, see Optimizer Hints. (Bug #36536936)
• Some subselects were not handled correctly. (Bug #36421690)
• Errors relating to SET subqueries were not handled correctly. (Bug #36335695)
• An invalid DDL statement in certain cases was not always rejected as expected. (Bug #35721121)
• Improved the internal function append_identifier(). (Bug #35633084)
• Normally, a view with an unused window definition should be updatable, but when it contained a subquery, it was marked as not updatable. At update time, the window was eliminated, but this was too late to allow an update to be performed.
• We fix this by testing mergeability, by checking the presence of window functions, rather than that of window definitions; this allows the view to be updatable, and the problematic UPDATE to succeed. (Bug #35507777)
• In some cases, SET did not perform correctly in prepared statements. (Bug #35308309)
• PARTITION BY ... (DEFAULT (column)) was not always handled correctly. (Bug #35044654)
• References: This issue is a regression of: Bug #33142135.
• This fix addresses the following issues:
  • Query_expression::is_set_operation() was not always executed properly.
  • Some sequences of DML statements could lead to an unplanned exit.
  • Some nested subselects were not always handled correctly.
• (Bug #34361287, Bug #35889583, Bug #35996409, Bug #36404149, Bug #36863048, Bug #37611264)
• On Debian, dh_strip_nondeterminism is no longer executed on zip and gzip files within the packages. (Bug #33791880)
• Removed an issue relating to invalid UTF8 values. (Bug #27618273, Bug #37709687)
• Addressed an issue relating to an invalid identifier. (Bug #22958632, Bug #37709664)
• Corrected an uninitialized variable in sql/statement/protocol_local_v2.cc. (Bug #117541, Bug #37622633)
• The LPAD() function did not return the correct value when given an empty string enclosed in double quotes unless the string's length exceeded the specified length variable. (Bug #117227, Bug #37498117)
• The null-safe equality operator (<=>) showed unexpected behavior when comparing multiple columns (ROW values) containing NULL. Fixed by simplifying the implementation of the operator. (Bug #117168, Bug #37462769)
• A negative impact in performance was observed when using a multivalued index with ORDER BY DESC and LIMIT in a query, where the value specified by LIMIT was greater than the number of rows actually in the result. (Bug #117085, Bug #37436310)
• References: This issue is a regression of: Bug #104897, Bug #33334911.
• When using MAX() as a window function, it returned NULL for the first row within the window frame, despite data existing in the first row. This happened when the start of the window frame was defined using N FOLLOWING, and the frame was ordered by the same expression as the argument of MAX(), possibly differing only in syntax such as aliases or table references, in descending order. A sequence of statements demonstrating the issue is shown here:

CREATE TABLE t0 (c0 INT);
INSERT INTO t0 VALUES (1), (2);
SELECT c0, MAX(c0) OVER (ORDER BY c0 DESC ROWS BETWEEN 1 FOLLOWING AND 1 FOLLOWING) FROM t0;

• We fix this by making sure that the first row number in the frame is set in the appropriate place in the program logic. (Bug #117013, Bug #37466984)
• Removed a double space within the INSERT IGNORE statements generated by mysqldump.
• Our thanks to Pieter Oliver for the contribution. (Bug #116845, Bug #37353658)
• The types of all ACL variables used internally have been changed to Access_bitmask.
• Our thanks to Mike Wang for the contribution. (Bug #116737, Bug #37318159)
• An error in include/assert_grep.inc could lead to erroneous results from any file that included it.
• Out thanks to Ke Yu for the contribution. (Bug #116239, Bug #37105430, Bug #37675340)
• If one client session had a uncommitted transaction that caused a DROP TABLE statement in another client session to be blocked, a third client session hung when trying to issue a USE DATABASE statement. (Bug #115706, Bug #36892499)
• The maximum for ssl_session_cache_timeout was defined as 84600 rather than 86400, and 84600 was stated erroneously to be the length of the day in seconds, in sql/ssl_init_callback.cc.
• Our thanks to Pika Mander for the contribution. (Bug #115165, Bug #37354555)
• Removed a memory leak from the mysqldump client. (Bug #111793, Bug #35621833)
• Removed the unused InnoDB and NDB handlerton get_tablespace() method. (Bug #109443, Bug #34916556)

Original source

Changes in MySQL 9.2.0 (2025-01-21)

• Account Management Notes
• C API Notes
• Character Set Support
• Compilation Notes
• Component Notes
• Configuration Notes
• Deprecation and Removal Notes
• Doxygen Notes
• Firewall Notes
• Installation Notes
• JavaScript Programs
• MySQL Enterprise Notes
• Optimizer Notes
• Performance Schema Notes
• Pluggable Authentication
• SQL Syntax Notes
• Functionality Added or Changed
• Bugs Fixed

Account Management Notes

• The database cache was not flushed properly following execution of DROP USER. (Bug #37132323)
• This release adds the CREATE_SPATIAL_REFERENCE_SYSTEM privilege, which allows the user to execute any of the statements CREATE SPATIAL REFERENCE SYSTEM, CREATE OR REPLACE SPATIAL REFERENCE SYSTEM, and DROP SPATIAL REFERENCE SYSTEM. Use of the SUPER privilege for this purpose should now be considered deprecated. For more information, see Privileges Provided by MySQL. (Bug #37046126, WL #16547)
• Failed password validation was not always handled correctly. (Bug #37041439)

C API Notes

• Process memory usage grew when the libmysqlclient API user tried to cache and reuse a prepared statement, preparing it once and then calling either of mysql_stmt_bind_param() or mysql_stmt_bind_named_param() followed by mysql_stmt_execute(), repeatedly without calling the matching mysql_stmt_close(), or calling it in the distant future (on application exit, for example).
• We solve this by introducing a separate MEM_ROOT object for storing the bind parameters array, which object can be cleared (deallocating the memory) on each call to mysql_stmt_bind_param() or mysql_stmt_bind_named_param(). (Bug #37202066)
• The Doxygen comments for the OK_Packet did not cover all possible cases; this has been updated so that it now accurately reflects the data actually sent by the server. (Bug #35630063)
• References: See also: Bug #35358417, Bug #36374232.

Character Set Support

• Zero rows resulted (where one row was expected) when selecting from a view created with its connection and client character set to latin1, where a query on the view used utf8 as its connection and client character set, the view contained literal values with non-ASCII characters, and the query performed a condition pushdown into a UNION of query blocks of the view.
• This problem was related to a previous issue which fixed an error for a similar problem: The problem in that case was properly considering the character set of the view definition when pushing down conditions contained in the view to inner query blocks, but the fix implemented at that time did not take into account the possibility that the view might contain non-ASCII characters.
• This meant that the condition to be pushed down was written to a text string with the wrong character set. We fix this oversight by ensuring that the string is created with the correct character set. (Bug #37111452)
• References: See also: Bug #36246859.

Compilation Notes

• macOS: Use OBJECT library in place of routing_guidelines when compiling with Xcode. (Bug #37350937)
• macOS: Removed obsolete CMake code from MacOS builds. (Bug #37258036)
• macOS: It is now possible to build MySQL using the Homebrew version of Clang. (Bug #37256912)
• macOS: Removed warnings of the form ld: warning: ignoring duplicate libraries and warnings specific to xcodebuild. (Bug #37065301)
• Microsoft Windows: Windows builds are now configured with -DWITH_ASAN. (Bug #37309813)
• Microsoft Windows: Disabled windows compiler warnings C26445 and C26821 in cmake/msvc_cppcheck.cmake. Both of these relate to MSVC substitution of gsl::span for std::span, which is not used for MySQL. (Bug #37158156)
• Microsoft Windows: Removed compiler warnings issued when building MySQL with Visual Studio 2022. (Bug #113870, Bug #36256477)
• Microsoft Windows: The CMake option -DWITH_SASL is not supported for building MySQL on Windows. Attempting to use this option when doing so now causes the build to terminate with an error. (Bug #113558, Bug #36155972)
• Solaris: The minimum required version of GCC to build MySQL on Solaris has been raised to 11.4. See Source Installation Prerequisites, for more information. (Bug #37256600)
• Starting with CMake 3.26, CMake writes the file CMakeFiles/CMakeConfigureLog.yaml, which supersedes CMakeError.log. References to CMakeError.log have therefore been removed. (Bug #37305289)
• Implemented the standards-compliant my_char_traits for use as a drop-in replacement for std::char_traits, which was deprecated in Clang 18 and removed in Clang 19. (Bug #37273525, Bug #37785339)
• The files stream_cipher.h and stream_cipher.cc, used by mysys/, were located in sql/ but did not depend on any other server code in this directory; these files have now been moved into mysys/. (Bug #37257736)
• Removed or fixed issues in a number of copyright header files. (Bug #37238155)
• Removed a maybe-uninitialized error found in sql/item.cc when building MySQL with GCC 14. (Bug #37157201)
• Disabled several clang-tidy checks which generated unnecessary warnings when compiling MySQL. (Bug #37129808)
• The CMake code for building component_mle assumed that GraalVM binaries were available, either under /usr/global/share, or in a location provided on the command line. Now, when neither of these conditions hold, it is possible to download GraalVM from a remote server. (Bug #37121798)
• The version of libedit used to compile MySQL was upgraded to 20240808-3.1. (Bug #37101293)
• The server could not be built on Ubuntu 22.04 using Clang 13. (Bug #37075154)
• Removed an error in mysql_prepare_create_table() (in the file sql/sql_table.cc) found when compiling MySQL with XCode 16. (Bug #37068527)
• Some plugins failed to load when MySQL was built with -FTLS-MODEL=initial-exec. Fixed by removing this compiler flag, and falling back to the default (global-dynamic) instead. (Bug #37017999)
• MySQL could not be compiled using Clang 19. (Bug #37014761)
• #include was missing from plugin/group_replication/libmysqlgcs/src/bindings/xcom/xcom/task.cc even though std::chrono::duration_cast() was referenced in this file. (Bug #116779, Bug #37329617)
• The server could not be built on Fedora 40 (and possibly other Linux platforms) using cmake 3.11, due to an issue with TIRPC. (Bug #116164, Bug #37080195)

Component Notes

• Group Replication: This release includes the Group Replication resource Manager component for MySQL Enterprise Edition, which monitors applier channel lag, recovery channel lag, and system resource usage on each secondary, and ejects any member whose lag or memory usage exceeds a user-configurable limit. Expelled servers may attempt to rejoin the group, provided that group_replication_autorejoin_tries is not set to 0.
• The operator can determine upper bounds for applier channel lag, recovery channel lag, and memory usage by setting the system variables group_replication_resource_manager.applier_channel_lag, group_replication_resource_manager.recovery_channel_lag, and group_replication_resource_manager.memory_used_limit, respectively.
• The Group Replication Resource Manager component checks the status of each secondary every 5 seconds, and updates a number of status variables which it provides, showing current channel lag and memory usage, as well as how many times a given threshold has been hit, and if the secondary has ever been expelled from the group, when this last occurred.
• See Group Replication Resource Manager Component, for more information. (WL #14940)
• For the MLE component, the mysql_option.option_usage table's USAGE_DATA column showed the used value as a string rather than as a boolean value.
• For more information, see Option Tracker Tables. (Bug #37122749)
• The entries present in the mysql_option.option_usage table depended on whether the Option Tracker was installed before or after the Audit Log.
• See Option Tracker Component, and MySQL Enterprise Audit Plugin (Deprecated), for more information. (Bug #37037438)
• Installing the Option Tracker component in one session while executing UNINSTALL COMPONENT in a different session caused the server to hang.
• See Option Tracker Component, for more information. (Bug #36991399)
• Errors in configuration files for keyring components were not properly logged. (Bug #36982002)
• INSTALL COMPONENT issued concurrently with a SET PERSIST which used a subquery could sometimes lead to an unplanned exit of the server. (Bug #36559078)
• References: See also: Bug #35647759.
• This release adds a Connection Control component (component_connection_control) intended to replace the Connection Control plugins (see Connection Control Plugins). A single component takes the place of both plugins, which are now deprecated and subject to removal in a future version of MySQL.
• One component replaces both of the deprecated Connection Control plugins; installation of the component requires a single INSTALL COMPONENT statement. See Connection Control Component Installation.
• The system and status variables associated with the plugin are also deprecated; see Connection Control Component Configuration for information about the variables supported by the Connection Control component in place of the plugin variables.
• In addition, the Information Schema CONNECTION_CONTROL_FAILED_LOGIN_ATTEMPTS table used by the plugins is now deprecated; the component keeps counts of failed connection attempts in a Performance Schema table connection_control_failed_login_attempts.
• See The Connection Control Component, for more information. (WL #16573)

Configuration Notes

• Microsoft Windows: On Windows, MySQL Configurator no longer executes the deprecated FLUSH PRIVILEGES statement. (Bug #37170330)

Deprecation and Removal Notes

• The FLUSH PRIVILEGES statement is now deprecated, and causes a warning when issued. You should expect this statement to be removed in a future MySQL release.
• The following constructs are also deprecated, and now cause a warning when used:
  • The FLUSH_PRIVILEGES privilege, and granting of this privilege
  • mysqladmin flush-privileges
  • mysqladmin reload.
• The features listed here do not cause any warnings but should be considered deprecated:
  • Flushing of privileges by SIGHUP
  • Flushing of the caching_sha2 cache by FLUSH PRIVILEGES
  • Flushing of privileges by mysqladmin refresh
• For more information, see FLUSH Statement, mysqladmin — A MySQL Server Administration Program, and Privileges Provided by MySQL. (WL #16567)
• The Version Tokens plugin is now deprecated, and subject to removal in a future MySQL release. The following related features are also now deprecated, and raise deprecation warnings whenever they are invoked:
  • The functions version_tokens_delete(), version_tokens_edit(), version_tokens_lock_exclusive(), version_tokens_lock_shared(), version_tokens_set(), version_tokens_show(), and version_tokens_unlock()
  • The VERSION_TOKEN_ADMIN privilege
  • The version_tokens_session and version_tokens_session_number server system variables
• Attempting to install the version_tokens plugin—or to start the server when this plugin is installed—also causes a deprecation warning to be issued. (WL #16571)

Doxygen Notes

• The payload length for COM_QUERY packets was calculated incorrectly in one of the MySQL packet examples. The payload length is now computed accurately, based on the actual length of the query.
• Our thanks to Konno Satoshi for the contribution. (Bug #116339, Bug #37161043)

Firewall Notes

• In some cases, after performing an upgrade, stored procedures relating to MySQL Enterprise Firewall were not processed correctly. (Bug #36084822)

Installation Notes

• When upgrading from MySQL 5.7 to a later MySQL release series, the system-created mysql.sys and mysql.session accounts are now modified to use the caching_sha2_password authentication plugin instead of the mysql_native_password plugin, which is deprecated in MySQL 8.0, and removed in MySQL 9.0. (Bug #36608636)

JavaScript Programs

• The MySQL ENUM and SET types are now supported for arguments of JavaScript stored routines. For more detailed information, including rules for conversion between these MySQL types and JavaScript types, see Conversion to and from MySQL ENUM and SET. (WL #16599)
• JavaScript programs supported by the MLE Component, available as part of MySQL Enterprise Edition, now supports access from JavaScript routines to user-defined functions, procedures, and variables. MySQL stored functions and procedures can be accessed using the Schema methods getFunction() and getProcedure() which return Function objects which can be used with arguments to invoke the routines. OUT and INOUT parameters of a stored procedure use placeholders created with mysql.arg(). For additional information and examples, see Stored Routine API.
• The MLE component also now provides a JavaScript MySQL transaction API which performs the actions of most MySQL transactional SQL statements, such as START TRANSACTION, COMMIT, ROLLBACK, and SET AUTOCOMMIT. Support for savepoints is also included. Support for an SqlError object is included. For more information, see JavaScript Transaction API, and SqlError Object.
• MySQL user variables can be accessed directly as properties of the JavaScript global Session object. For example, a user variable named myvar can be read or set as session.myvar. See Accessing Session Variables from JavaScript, for more information and examples.
• This release also adds support for direct access to the MySQL built-in functions RAND(), SLEEP(), UUID(), and IS_UUID(), as, respectively, rand(), sleep(), uuid(), and isUUID(). See MySQL Functions.
• For further information about JavaScript programs and the MLE Component, see JavaScript Stored Programs, and Multilingual Engine Component (MLE). (WL #16585)
• The MLE component, available as part of MySQL Enterprise Edition, now supports reusable JavaScript libraries containing functions which can be called from other JavaScript stored programs. JavaScript libraries can be managed using the CREATE LIBRARY and DROP LIBRARY SQL statements added in this release; they can be included in other stored JavaScript programs with the USING clause added to CREATE FUNCTION and CREATE PROCEDURE; USING supports a list of one or more library names.
• CREATE LIBRARY creates a new JavaScript library in a given database. Library code is parsed and checked for validity at creation time, and is rejected if it contains errors.
• DROP LIBRARY drops a JavaScript library. Library functions can be referred to in other JavaScript stored programs using library_name.function_name notation. Libraries can be aliased with USING when including them in JavaScript programs.
• You can obtain the code from a library using the SHOW CREATE LIBRARY statement, also new in this release. Two new related Information Schema tables have also been added: The LIBRARIES table provides information about JavaScript libraries, and the ROUTINE_LIBRARIES provides information about stored routines which include JavaScript libraries.
• Counts of library DDL and SHOW CREATE LIBRARY statements which have been issued on the server are available as the status variables Com_create_library, Com_drop_library, and Com_show_create_library.
• For more information and examples, see Using JavaScript Libraries, as well as Multilingual Engine Component (MLE). (WL #16359, WL #16360, WL #16362, WL #16555)

MySQL Enterprise Notes

• Replication: The rnd_pos() function of the handler interface for replication applier metrics tables was implemented wrongly; in cases where this function was used, the tables appeared to be missing one row.
• For more information, see Replication Applier Metrics Component. (Bug #37076428, Bug #37132660)
• Replication: For most wait operations, we note the time the wait started and the time it ends, and add their difference to the total waiting time, but this was not the case when tracking time spent waiting for the commit order, where we used a function that woke up every second and caused the wait time to be incremented by 1. This 1-second precision was too low for most practical purposes and caused unnecessary complication of the code.
• To fix these issues, we now track waiting time in the same way for commit order waits as we do for other waits.
• For more information, see Replication Applier Metrics Component. (Bug #37053708)
• Replication: The MySQL Option Tracker component can now provide usage information about the binary log, group replication, and use of the server as a replica whenever these features are enabled.
• For more information, see Option Tracker Supported Components. (WL #16529)

Optimizer Notes

• JSON: Added lookup references to iterator-based EXPLAIN FORMAT=JSON for index lookups. The lookup_references field in the JSON v2 EXPLAIN format corresponds to the ref field in the JSON v1 EXPLAIN format. (Bug #37126176)
• Some hash joins spent an unreasonably high time in pack_rows::RequestRowId(), even though they did not use row IDs.
• Fixed by skipping the loop over TableCollection::tables() in RequestRowId() in the case where there are no tables from which to request row IDs. (Bug #37243461)
• EXPLAIN FORMAT=TREE now shows the clustered primary key scan for RowID-Ordered Retrieval (ROR) intersection plans. (Bug #37199800)
• Pushing down a condition which had an aggregate function in a WHERE clause caused the aggregate function to be evaluated when it should not have been. (Bug #36421735)

Performance Schema Notes

• If a user other than root ran START REPLICA, PERFORMANCE_SCHEMA.PROCESSLIST assigned that user's name to the newly spawned foreground replication threads instead of system user.
• As of this release, system user is assigned to all foreground system threads. (Bug #37357562)
• PERFORMANCE_SCHEMA.PROCESSLIST filtered foreground threads which did not have a user name. As of this release, system_user is assigned to foreground threads without a user name. (Bug #37357562)
• Under certain circumstances, a metadata lock can be upgraded or downgraded to a different LOCK_TYPE. This change was not reflected in the PERFORMANCE_SCHEMA.METADATA_LOCKS table.
• Our thanks to George Ma and the Alibaba team for the contribution. (Bug #116625, Bug #37271768)

Pluggable Authentication

• The following issues relating to the AUTHENTICATION_POLICY_ADMIN privilege have now been resolved:
  • For a user having the privilege, it was not possible to create a user for whom a default authentication plugin was omitted, because the defaults specified in the global value of authentication_policy were included at creation time whether any were specified or not.
  • We fix this by ignoring the global authentication policy when the user performing account creation has the AUTHENTICATION_POLICY_ADMIN privilege.
  • For a user not having the privilege, an attempt to create a user without specifying any authentication factors was rejected with an error, because mandatory factors in the default global authentication policy were not included when they were not specified.
  • We fix this by including any mandatory factors from the global authentication policy when the user performing the account creation does not have the AUTHENTICATION_POLICY_ADMIN privilege.
• For more information, see the description of the authentication_policy system variable, as well as Privileges Provided by MySQL. (Bug #37027739)
• The authentication_openid_connect plugin is now registered with the Option Tracker component.
• For more information, see Option Tracker Component, and OpenID Connect Pluggable Authentication. (Bug #116045, Bug #37041216)

SQL Syntax Notes

• Important Change: The BINLOG keyword can no longer be used as an unquoted label name in MySQL stored programs. Prior to upgrading to this release, you should update any affected applications accordingly.
• For more information, see Keywords and Reserved Words. (Bug #22574003)

Functionality Added or Changed

• InnoDB: During recovery, IBUF merges were disabled for all recovery batches except for the last batch, but is now disabled for all batches. This change prevents the addition of new redo log entries during recovery, as they led to deadlock and recovery performance degradation with reads and writes from the additional IBUF pages. (Bug #33134355, WL #15372)
• Microsoft Windows: On Windows, MySQL Configurator has added a command-line interface for configuring a MySQL Server installation. See MySQL Configurator CLI, for more information.
• A known limitation of this release: the new CLI only supports the configure action. (WL #16564)
• Binary packages that include curl rather than linking to the system curl library have been upgraded to use curl 8.11.1. (Bug #37389565)
• Refactored code to take advantage of OpenSSL 3.x functionality to improve performance for cryptographic operations, such as those performed by the MySQL MD5() and SHA2() functions. (Bug #116939, Bug #37185170, WL #16504)
• Added a source code static analysis tool. For usage details, see scripts/static_analysis.md in the source code or execute python3 ./scripts/static_analysis.py --help for additional help. Use this to check single commits or the entire source code repository. (WL #16329)
• Converted the caching SHA-2 authentication plugin to use the event component API instead of the audit log plugin API. This also removes the sha2_cache_cleaner audit plugin. (WL #16572)

Bugs Fixed

• Incompatible Change: Corruption occurred in a spatial index when an update of a geometry with a minimal change in the minimum bounding rectangle (MBR) was followed by a delete operation.
• When upgrading to this release, it is recommended that you drop any spatial indexes beforehand, then re-create them after the upgrade is complete. Alternatively, you can drop and re-create such indexes immediately following the upgrade, but before making use of any of the tables in which they occur. You should also be aware that downgrading to any previous version reintroduces the original problem described previously.
• For more information, see Creating Spatial Indexes. (Bug #36452528)
• InnoDB: Improved the simulated asynchronous I/O (AIO) handler performance for high volume situations. (Bug #37366607)
• InnoDB: Improved asynchronous I/O (AIO) page cleaner thread management performance. (Bug #37359213)
• InnoDB: Concurrently truncating a table while querying the Performance Schema sometimes cause MySQL to halt unexpectedly. (Bug #37271715)
• InnoDB: It was possible for an ALTER TABLE operation using the INPLACE algorithm on a table containing both a spatial index and an auto-increment column to cause corruption or, in debug builds, to trigger a debug assert. This was due to the auto-increment column value being overwritten in the old records of the spatial index while the new record was prepared. (Bug #37189985)
• InnoDB: Certain IO buffer serializations triggered an assertion in debug builds that caused the system to hang. (Bug #37139618)
• InnoDB: Improved InnoDB start up time. (Bug #36880863)
• References: This issue is a regression of: Bug #36808732.
• InnoDB: An assertion failure was raised when creating a FULLTEXT index on a table with an FTS_DOC_ID value greater than 4294967295. (Bug #36879147)
• References: See also: Bug #37387224.
• InnoDB: Improved can_older_trx_be_still_active() so that it no longer uses trx_sys->rw_trx_list. (Bug #36729529)
• InnoDB: Dropping a primary key, and then adding a new AUTO_INCREMENT column as a primary key in descending order using the INPLACE algorithm failed.
• Our thanks to Shaohua Wang and the team at Alibaba for the contribution. (Bug #36658450)
• InnoDB: Removed a memory leak from the mysqladmin client. (Bug #36537941)
• InnoDB: Extending a user tablespace produces file extension redo log records (MLOG_FILE_EXTEND), but they were not produced when extending the system tablespace. (Bug #36511673)
• InnoDB: On Windows, fixed a doublewrite buffer regression that slowed file access, and refactored FILE_FLAG_OVERLAPPED flag usage for opening files. (Bug #36259487)
• InnoDB: A DELETE operation on a table with a self referential foreign key and full-text index could have triggered an assertion. (Bug #36234681)
• InnoDB: Removed code that handled obsolete redo log formats. (Bug #35020216)
• InnoDB: Common prefix compression for redo log inserts (MLOG_REC_INSERT) was disabled but is now enabled when the versions match. (Bug #34946626)
• References: This issue is a regression of: Bug #13899.
• InnoDB: Virtual column information for a row containing an externally stored BLOB was not always logged during an UPDATE operation, which sometimes resulted in an Index PRIMARY is corrupted error. (Bug #34574604)
• InnoDB: ON DELETE CASCADE with generated columns containing secondary indexes sometimes failed, due to virtual column templates not being initialized before deletion.
• Our thanks to Rahul Malik for the contribution. (Bug #33691659)
• InnoDB: An update operation attempted to update a virtual column while building an update node for a child table, but should not have done so since foreign key constraints cannot reference virtual columns. (Bug #33327093)
• InnoDB: If binary logging was disabled, it was possible for full-text search queries to be incorrect after a server crash.
• Our thanks to Yin Peng and the Tencent team for the contribution. (Bug #116212, Bug #37095383)
• InnoDB: Removed a duplicate declaration of the dd_set_autoinc() method. (Bug #116175, Bug #37089340)
• InnoDB: It was possible for ALTER TABLE, which rebuilds InnoDB tables using the INPLACE algorithm, to be rejected with a duplicate key error due to a non-duplicate record being inserted concurrently while the rebuild was paused to release a page latch.
• Our thanks to Dmitry Lenev and the team at Percona for contributing to this fix. (Bug #115511, Bug #36808088)
• InnoDB: The check enforcing the rule that ALGORITHM=INSTANT cannot be used on a column referenced by a foreign key constraint from another table did not inspect the last field of said constraint. (Bug #115457, Bug #36796094)
• InnoDB: CPU usage statistics did not account for a processor count over 128, which could degrade performance on these larger systems. (Bug #115399, Bug #36765223)
• InnoDB: Executing ALTER TABLE with ADD COLUMN or DROP COLUMN against an empty table now uses the INPLACE algorithm by default instead of INSTANT. This change means the row version is no longer incremented for these simple operations. (Bug #113051, Bug #36004394)
• InnoDB: An ALTER TABLE operation that rebuilt an InnoDB table using the INPLACE algorithm potentially led to losing a row of data if a purge occurred concurrently on the altered table that contained a clustered or spatial index.
• Our thanks to Dmitry Lenev and the team at Percona for contributing to this fix. (Bug #110706, Bug #113812, Bug #115608, Bug #116764, Bug #35303494, Bug #36261480, Bug #36846567, Bug #37318367)
• InnoDB: Queries with a descending primary key and the index_merge optimization sometimes yielded incorrect results such as missing rows. (Bug #106207, Bug #33767814)
• Replication: In an InnoDB ClusterSet setup, when autocommit was set to OFF on all nodes in the cluster, a controlled switchover using MySQL Shell was rejected with Error 1105 (Unknown error).
• To fix this, we now force a new transaction in channel_change_source_connection_auto_failover() whenever autocommit=OFF to prevent table access deadlocks when an info repository transaction is executed after changing SOURCE_CONNECTION_AUTO_FAILOVER. (Bug #37173907)
• Replication: While large transactions were being received and applied, and a request to stop the replication channel was made using STOP REPLICA, MySQL did not do so properly, and subsequently did not process any channel commands. In addition, the server shutdown process did not complete gracefully, and required either the MySQL process to be killed or the host system to be restarted. (Bug #115966, Bug #37008345)
• Replication: The log message written when a replica reconnects to the source (when, for example, it is stopped and restarted by issuing STOP REPLICA followed by START REPLICA) , While initializing dump thread for replica with UUID uuid, found a zombie dump thread with the same UUID. Source is killing the zombie dump thread(thread_id) has been improved to Upon reconnection with the replica, while initializing the dump thread for UUID uuid, an existing dump thread with the same UUID was detected. The source is terminating the previous dump thread (thread_id), which is normal and expected. (Bug #84358, Bug #25330090)
• Group Replication: Removed a potential race condition between the internal functions cs::apply::Commit_order_queue::front() and cs::apply::Commit_order_queue::remove(). (Bug #37223451)
• References: See also: Bug #35206392.
• Group Replication: When the primary node unexpectedly left the group and quickly attempted to rejoin, the member which had been elected to remove other, faulty members tried to expel or remove the faulty node but could not do so due to lack of a majority. When the expelled or removed node was the primary, this left the cluster without a primary, resulting in an unusable state. (Bug #36991859)
• References: See also: Bug #37181867.
• Group Replication: In some cases, adding a new secondary caused existing secondaries to lag, leading to a deadlock which persisted with no more writes possible until the primary was restarted.
• This deadlock occurred between the ticket manager, which ensures that transactions are committed on the correct side of a view change (before or after the view change), and the commit order manager on the inbound replication channel, which ensures that transactions are committed in the same order in which they are received, when these two managers required different orders. This meant that, in some cases, adding a new secondary caused the group primary to be unable to do writes.
• We solve this issue by ignoring the commit order manager ordering and enforcing the ticket manager ordering for non-conflicting transactions when such a deadlock occurs. A consequence of this is that replica_preserve_commit_order may not be strictly honored near a View_change_log_event. In other words, replica_preserve_commit_order no longer provides a strict guarantee on an inbound channel on a Group Replication primary. replica_preserve_commit_order still ensures that transactions are ordered correctly, with the only exception being non-conflicting transactions around view change log events. (Bug #35206392)
• References: See also: Bug #37223451.
• Group Replication: Improved garbage collection in Group Replication by eliminating non-essential calls to is_subset_not_equals(). (Bug #110673, Bug #35286974)
• Group Replication: Removing a group member from a group in which all members were running the same version of MySQL, upgrading it to a later version (from a later release series), and then requesting it to rejoin the group caused the upgraded group member to hang in the recovering state.
• Improved an existing fix made in MySQL 9.1.0 for a condition which allowed queries with unknown columns in scope. (Bug #37341055)
• References: This issue is a regression of: Bug #35058815.
• Calling mle_set_session_state() with a type argument that was not a JSON string caused the component to exit. (Bug #37334566)
• A subquery which was marked for materialization during resolution was not optimized when an impossible condition was detected else where in the query. This created problems when optimizer was trying to detect subqueries that needed to be materialized. Since the subquery was not optimized, trying to access the query block's join pointer resulted in a server exit.
• We solve this by checking to see whether the subquery was optimized before finalizing the materialization transform for it. (Bug #37321762, Bug #37847157)
• Certain nested SELECT statements produced ER_DUP_KEY on a temporary table. (Bug #37309915)
• References: This issue is a regression of: Bug #115597, Bug #36846564.
• A query having a subquery which did not use any tables and which had a nonzero OFFSET clause did not return the correct result. For example, SELECT (SELECT 1 LIMIT 1 OFFSET 10) returned 1 instead of the expected NULL. (Bug #37293822)
• The errors ER_DD_UPDATE_DATADIR_FLAG_FAIL, ER_IB_MSG_FIL_STATE_MOVED_PREV_OR_HAS_DATADIR, ER_RPL_KILL_OLD_DUMP_THREAD_ENCOUNTERED, and ER_RPL_MTA_ALLOW_COMMIT_OUT_OF_ORDER were originally defined in MySQL 8.0, but were subsequently assigned different error code numbers (but with the same names) in MySQL 8.4. The numbers assigned in MySQL 8.0 now apply to MySQL 8.0 only; in the MySQL 8.4 and later release series, only the numbers assigned in MySQL 8.4 are used. (Bug #37284176)
• Added missing checks for NULL input arguments in mle::validate() and validate_session_options(). (Bug #37267887)
• For hash joins, time and resources were spent in unnecessary construction and destruction of HashJoinChunk objects that were never used, even in cases where a hash join could not be performed completely in memory and subsequently spilled to disk. (Bug #37235166)
• Adding an event to the sys schema during an upgrade led to an unplanned shutdown of the server. (Bug #37162611)
• In sql/item_cmpfunc.cc, Item_bool_func2::resolve_type() made an unchecked call to Item_bool_func::resolve_type(); the call to Item_bool_func::resolve_type() ignored its return value, and execution continued even in case of an error. (Bug #37143289)
• When evaluating an expression such as value BETWEEN 'a' AND 'b', the server now checks the string literals and ensures that they use the same character set and collation as the value to be compared with them. (Bug #37086818)
• Removed an unnecessary assertion in sql/item_func.cc. (Bug #37083848)
• References: See also: Bug #29467577.
• Removed a memory leak found in option_tracker/udfs.cc. (Bug #37075241)
• AppArmor denied access to /proc/$pid/task/$thread_id/mem, a file required to generate a stack trace. (Bug #37063288)
• References: See also: Bug #37387034.
• Added a missing check for an empty pointer to WalkAndReplace(), in sql/sql_resolver.cc. (Bug #36987582)
• References: This issue is a regression of: Bug #112557, Bug #35855294.
• MySQL allowed outer references in window PARTITION BY and ORDER BY expressions, which was not in accordance with the SQL standard. (Bug #36921175, Bug #37847161)
• A filter condition in a subquery was sometimes ignored when a query used the index_subquery join type for subquery execution, and the subquery table used materialization in the execution plan. The derived table access path replaced the filter condition, resulting a final plan without the filter layer. To fix this, in such cases, we now add the derived table access path along with the filter access path instead of replacing the latter. (Bug #36918913)
• Transformations of some scalar subqueries to derived tables were not always performed correctly. (Bug #36902116)
• Some UNION operations similar to a UNION b UNION c ... consumed excessive memory. To help keep this from happening, we now flatten equal set operations at the parsing level, before contextualization occurs, which should result in reduced resource usage by such operations. (Bug #36652610)
• Improved the internal function my_print_help(). (Bug #36615714)
• References: See also: Bug #37387224.
• Removed incorrect code from Acl_cache. (Bug #36608160)
• A subquery containing an aggregate function WITH ROLLUP which was part of a row value comparator was not always processed correctly. (Bug #36593235)
• References: See also: Bug #37387180. This issue is a regression of: Bug #30969045, Bug #30921780, Bug #26227613, Bug #29134467, Bug #30967158.
• It was possible for errors raised when persisting variables not to be reported correctly. (Bug #36574732)
• Some subqueries using WITH ROLLUP were not always processed correctly. (Bug #36421704)
• MyISAM block length calculations were not always performed correctly. (Bug #36347992)
• Removed a potential race condition. (Bug #35981769)
• References: See also: Bug #36608160.
• The mysqldump --column-statistics option is now disabled by default. (Bug #35209008)
• Fixed an issue relating to FTS and concurrent DDL or DML. (Bug #34633727)
• DROP VIEW name was rejected with ER_BAD_TABLE_ERROR if there existed a table with the same name. (Bug #33200087)
• Setting explain_json_format_version to 2 now has the following effects on the output of EXPLAIN FORMAT=JSON:
  • The output includes the JSON schema version in major.minor format; this is always displayed as 2.0 in MySQL 9.2.
  • Only query attributes are stored directly in the top-level object; in version 1 output, this object contains both query attributes and iterator attributes.
• The output of EXPLAIN FORMAT=JSON when explain_json_format_version = 1 remains completely unchanged in this release. (Bug #116915, Bug #37372130, Bug #35239659)
• The mysqlslap utility did not disable SSL when the --ssl-mode=disable option was passed in. (Bug #116844, Bug #37353662)
• When taking the UNION of a YEAR column and a BOOL (TINYINT) column, the data type of the result was TINYINT, due to a flaw in the internal field_types_merge_rules array. We fix this by adjusting the result data type in such cases to SMALLINT. (Bug #116415, Bug #37192491)
• Incorrect results were returned by some queries that used hash antijoins when the hash table did not fit in the join buffer and spilled to disk. (The query triggering the issue actually specified LEFT JOIN, but this was transformed internally from a left outer join to an antijoin.)
• The problem was that some rows in the probe table were skipped when writing the probe rows to chunk files, the skipped rows being those that had NULL in part of the join key. Such rows can be skipped for inner joins and semijoins, as they are known to have no match in the build table, but for outer joins and antijoins, rows in the probe table which have no matching row in the build table should be part of the join result, so they must be included in the chunk files.
• We already preserved these rows in the chunk files for outer joins. This fix extends the logic used for that purpose so that it also applies for antijoins. (Bug #116334, Bug #37161583)
• A query containing a common table expression produced different results on the second and subsequent executions in some cases when used in a prepared statement or stored procedure. (Bug #116140, Bug #37074489)
• When the server sql_mode included ANSI_QUOTES, SHOW GRANTS quoted stored function and stored procedure names in backticks (`), while using double quotes (") with the names of other objects. Now double quotes are always used in such cases. (Bug #115953, Bug #37003502)
• Window functions having BIT values as arguments did not return BIT. (Bug #115597, Bug #36846564)
• The mysqlslap utility became unresponsive if --auto-generate-sql-execute-number and --concurrency were specified without --auto-generate-sql or --query. (Bug #113215, Bug #36048754)
• In MySQL 8.0 and later, queries of the form SELECT DISTINCT ... FROM t1 WHERE NOT IN(SELECT ...) were transformed into an antijoin if possible, causing the optimizer not to choose a group skip scan for table t1 whereas it would have been chosen in MySQL 5.7. This resulted in a performance degradation for such queries. Group skip scan is not chosen, since the query is now no longer a single-table query following the antijoin transformation, and this access method is enabled only for single table queries. The same behavior can be seen for queries which are transformed into semijoins as well. In such cases, group skip scan access method can still be used if the access method is used only for duplicate removal (that is, with DISTINCT or GROUP BY, but without aggregate functions).
• To fix this, we enable group skip scan when there is only one table in the original query, irrespective of the number of semijoin tables present after internal transformations as long as the query contains no aggregate functions. (Bug #112362, Bug #35842412)
• The mysql client did not allow using '#' or '--' inside an optimizer hint comment.
• Our thanks to Kaiwang Chen for the contribution. (Bug #98521, Bug #30875669)
• The mysqldump --compact option now disables --tz-utc.
• Previously, with --tz-utc enabled (the default), --compact executed SET TIME_ZONE='+00:00' before reading data but did not write this statement to the dump file. This caused an inconsistency, since data was extracted using UTC but readers of the dump did not know that the data used this time zone. (Bug #58491, Bug #11765514)
• When comparing internal representations of VECTOR columns, their length properties were not considered.

Original source

Changes in MySQL 9.1.0 (2024-10-15)

• Atomic DDL Notes
• Audit Log Notes
• Authentication Notes
• C API Notes
• Compilation Notes
• Component Notes
• Configuration Notes
• Deprecation and Removal Notes
• SQL Function and Operator Notes
• JavaScript Programs
• Keyring Notes
• MySQL Enterprise Notes
• Performance Schema Notes
• Pluggable Authentication
• SQL Syntax Notes
• Thread Pool Notes
• Trigger Notes
• Vector Data Type
• Functionality Added or Changed
• Bugs Fixed

Atomic DDL Notes

Previously, CREATE DATABASE and DROP DATABASE were atomic but were not fully crash-safe. This could lead to issues such as leftover database directories without data dictionary records or incomplete rollbacks on DROP DATABASE failures. These statements are now fully atomic and crash-safe, provided all tables use a storage engine supporting atomic DDL (e.g., InnoDB). See Atomic Data Definition Statement Support. (WL #16232)

Audit Log Notes

The audit log treated users with names not starting with alphanumeric characters (e.g., '$foo') as invalid, despite MySQL allowing such users. This also affected MySQL Enterprise Firewall. (Bug #36778917)

Authentication Notes

Microsoft Windows: MySQL webauthn plugins now support Windows Hello authentication with Windows 11+ using the authentication_webauthn plugin. The Windows Hello passkey store is accepted as a software “device”. The server plugin is available in MySQL Enterprise Edition; the client plugin is available in both Enterprise and Community Editions. A new mysql client option --plugin-authentication-webauthn-device=# allows selecting among multiple devices. Only keys stored in the Windows passkey store are supported. Passkey deletion must be done using system tools, not via the mysql client. See WebAuthn Pluggable Authentication. (WL #16293)

Added OpenID Connect support for MySQL Enterprise Edition using the authentication_openid_connect server-side plugin. See OpenID Connect Pluggable Authentication. (WL #16269)

C API Notes

The asynchronous interface used unsafe static local variables to store state information. (Bug #115703, Bug #36891894)

Compilation Notes

MySQL did not compile on Fedora 41 and Ubuntu 24.10. Added CONTRIBUTING.md and SECURITY.md files to sources. Aligned CMAKE_MINIMUM_REQUIRED with correct version and policies. Upgraded bundled curl to 8.9.1, zlib to 1.3.1, libcurl to 8.9.0, lz4 to 1.10.0, TI-RPC to 1.3.5, opentelemetry-cpp to 1.15.0. Added CMake code for Tesseract OCR library. Replaced custom MY_TARGET_LINK_OPTIONS with TARGET_LINK_OPTIONS. Fixed -DWITHOUT_SERVER option on Enterprise Linux 7. Dropped Fedora 28 and 34 support. Fixed linking with mold on Enterprise Linux 9. Added DISABLE_PERFSCHEMA build option. Upgraded libfido to 1.15.0. Bundled RE2 regex library. Upgraded clang-format from 10 to 15. Fixed SQL_I_list member initialization and assignment. Thanks to contributors. (Various Bug numbers)

Component Notes

Group Replication: Added component_group_replication_flow_control_stats for statistics about transactions throttled by flow control, available in MySQL Enterprise Edition. Requires Group Replication plugin installed. Provides global status variables for flow control throttling. See Group Replication Flow Control Statistics Component. (WL #16239)

SHA3-224 algorithm was used when SHA3-256 was specified as encryption method. (Bug #36980306)

mysql/plugin.h was included unnecessarily by some logging components. (Bug #34772742)

Option Tracker component added, implementing tables performance_schema.mysql_option and mysql_option.option_usage to provide information about server options and usage. Requires OPTION_TRACKER_UPDATER privilege. Available in MySQL Enterprise Edition. See Option Tracker Component and related documentation. (WL #16217, WL #16403, WL #16503)

Introduced Amazon Web Services keyring component (component_keyring_aws), superseding the deprecated AWS keyring plugin (keyring_aws). See Using the component_keyring_aws AWS Keyring Component and Migration from AWS keyring plugin. (WL #16348)

Configuration Notes

Microsoft Windows: Fixed MySQL Configurator revert button, data directory browsing outside ProgramData folder, unnecessary background operations after clicking Next if options unchanged, and empty password strength description. (Various Bug numbers)

Deprecation and Removal Notes

NDB Cluster APIs: Removed support for Node.js deprecated in NDB Cluster 9.0. (WL #16246)

keyring_aws plugin deprecated and subject to removal; users should migrate to AWS keyring component. See Using the component_keyring_aws AWS Keyring Component and Migration from AWS keyring plugin. (WL #16348)

SQL Function and Operator Notes

DATABASE() output was truncated when used in UNION queries. (Bug #36871927)

SUM() yielded different results when DISTINCT was used. Fixed by using already determined data type and length for temporary tables with DISTINCT. (Bug #115476, Bug #36796197)

JavaScript Programs

JavaScript stored programs always use utf8mb4 character set; COLLATION_CONNECTION column now shows utf8mb4_0900_ai_ci instead of latin1. This may affect comparisons using COLLATION_CONNECTION. See workaround with explicit collation. (Bug #36925906)

Fixed blob value invalidation after prepared statement deallocation inside stored routines. (Bug #36902680)

Fixed unplanned server exit when using user variables in DML within stored procedures using Statement handle interface. (Bug #36892945)

SqlResult now supports iterators for result sets. (Bug #36835486)

Multilingual Engine Component (MLE) emits Telemetry metrics. SQL queries in stored JavaScript procedures observable in Performance Schema query digests and events_statements_history_long. (Bug #36216014, WL #16354)

VECTOR type fully supported by JavaScript stored programs. See JavaScript Stored Program Data Types and Argument Handling. (WL #16405)

Keyring Notes

Removed support for insecure AES ECB mode from all MySQL keyring-related components. (Bug #36749277)

MySQL Enterprise Notes

Added MySQL Replication Applier Metrics component providing statistical information about replication in Performance Schema tables replication_applier_metrics and replication_applier_progress_by_worker. Enhances observability of replication. See Replication Applier Metrics Component. (WL #15620)

Performance Schema Notes

Fixed issues with OTLP exporter. (Bug #36792180, Bug #36783070)

Enabled command line/config file parameter performance-schema-meter to enable/disable Telemetry meters. See Server Meters. (Bug #36698082)

Fixed erroneous comment in storage/perfschema/table_host.h. Thanks to Sho Nakazono. (Bug #115844, Bug #36954266)

Improved performance of SELECT * FROM sys.innodb_lock_waits; by limiting locks fetched. Added primary keys to DATA_LOCKS and DATA_LOCK_WAITS. (Bug #100537, Bug #31763497)

OpenTelemetry Logging enables exporting telemetry logs to OpenTelemetry backends. Implemented as Telemetry Logging Component (Enterprise Edition and HeatWave) and Telemetry Logging Interface (Community, Enterprise, HeatWave). See Installing OpenTelemetry Support and MySQL Server Doxygen Documentation. (WL #15201, WL #15202)

Pluggable Authentication

Fixed SET PERSIST authentication_ldap_simple_bind_root_pwd= password not saving password. (Bug #35478833)

SQL Syntax Notes

Disallowed identifier chains referencing columns in derived tables that contradict SQL standard. (Bug #36704815)

Added IF NOT EXISTS clause for CREATE VIEW statement. If view exists, statement succeeds with warning and does not change view. IF NOT EXISTS incompatible with OR REPLACE clause. See CREATE VIEW Statement. (WL #14774)

Thread Pool Notes

Fixed handling of some concurrent connection attempts. (Bug #36625082)

Trigger Notes

Performance improvements by separating parsing/loading of triggers into metadata and execution phases. Added server variable table_open_cache_triggers to control eviction of tables with fully loaded triggers. Added status variables Table_open_cache_triggers_hits, Table_open_cache_triggers_misses, Table_open_cache_triggers_overflows. Improved memory usage for trigger errors. SHOW CREATE TRIGGER now displays complete trigger definitions even if bodies cannot be parsed. See How MySQL Opens and Closes Tables. Thanks to Dmitry Lenev. (Bug #44625, Bug #86821, Bug #11753220, Bug #26366333, WL #16455)

Vector Data Type

STRING_TO_VECTOR() function now allows trailing whitespaces in string representations of VECTOR arguments. Parser trims whitespace after numbers and brackets. Examples: "[1 ,2]", "[1,2 ]", " [1,2]", "[1,2] ". (Bug #37009633)

Functionality Added or Changed

Important Change; InnoDB: Default innodb_log_writer_threads is OFF on systems with fewer than 32 logical processors. (WL #16396)

Important Change: OpenSSL library updated to 3.0.15 where bundled. See OpenSSL 3.0 Series Release Notes and OpenSSL Security Advisory [3rd September 2024]. (Bug #37021075)

Performance; Replication: Changed binary log transaction dependency tracking data structure to ankerl::unordered_dense::map, reducing space by ~60%. (Bug #37008442, Bug #37529256)

Group Replication: Added INFO level logging messages for action begin, message blocking, and message receipt with internal counter decrement. (Bug #32929873)

Added spaces around = sign in lookup_condition for EXPLAIN FORMAT=JSON and EXPLAIN FORMAT=TREE outputs. (Bug #36825515)

EXPLAIN FORMAT=TREE output now includes semijoin strategy info. (Bug #36773414)

Added "multi_range_read": true to EXPLAIN FORMAT=JSON output when MRR is used. (Bug #36614948)

Added --system-command option to mysql client to enable/disable system client command. Disabled by default. (Bug #36377685, WL #16482)

MySQL server supports alternative syntax for GROUP BY ROLLUP modifier. See GROUP BY Modifiers and SELECT Statement. (WL #15843)

Added status variables TempTable_count_hit_max_ram and Count_hit_tmp_table_size to track internal temporary table conversions from memory to disk. (WL #16377)

Maximum TOTAL_ROW_VERSIONS value changed from 64 to 255 for ADD COLUMN and DROP COLUMN operations using INSTANT algorithm. (WL #16369)

Added server system and status variables for tracking memory usage by queries when global_connection_memory_tracking is enabled. Variables include global_connection_memory_status_limit, Count_hit_query_past_global_connection_memory_status_limit, connection_memory_status_limit, Count_hit_query_past_connection_memory_status_limit. These are for testing and do not cause query rejection. (WL #16097)

Bugs Fixed

Numerous bug fixes across InnoDB, Group Replication, Performance Schema, optimizer, stored procedures, triggers, and other components. Fixes include memory leaks, crashes, incorrect query results, compilation issues, and regressions. Notable fixes:

• InnoDB adaptive hash index inline function regression fixed. (Bug #111538)
• Fixed compilation on Fedora 41 and Ubuntu 24.10.
• Fixed memory leak in bulk loader. (Bug #35988311)
• Fixed optimizer issues with multi-valued virtual columns and FTS index optimization.
• Fixed server halts on DELETE/UPDATE with INSTANT algorithm.
• Fixed various Group Replication memory leaks and connection handling issues.
• Fixed incorrect results with DISTINCT and ORDER BY in derived tables.
• Fixed SQL syntax and semantic issues with GROUP BY, ROLLUP, UNION, and window functions.
• Fixed issues with prepared statements, triggers, and stored routines.
• Fixed OpenSSL and Kerberos library updates.
• Fixed issues with mysqldump escaping and output.
• Fixed various regressions and assertion failures.

(For full details, see the MySQL 9.1.0 Release Notes.)

Original source

Bugs Fixed

• InnoDB: In some cases, following the creation of a very large number of tables (8001 or more), the server could not be restarted successfully. (Bug #36808732)

References: This issue is a regression of: Bug #33398681.

• InnoDB: Improved tablespace file scan performance at startup. (Bug #110402, Bug #35200385)

• Group Replication: Running a CREATE TABLE ... SELECT statement on a source coming from an asynchronous channel to Group Replication led to errors on the replica. (Bug #36784284)

Original source

Important
This release is no longer available for download. It was removed due to a critical issue that could stop the server from restarting following the creation of a very large number of tables (8001 or more). Please upgrade to MySQL 9.0.1 instead.

Audit Log Notes
• The audit log plugin printed the warning Plugin audit_log reported: 'Cannot process audit log file. File name timestamp value is missing or invalid: ' dbname '' when no audit log file was being processed. (Bug #36400810)
• With large files (200 MB or larger), executing audit_log_read() took an excessive amount of time to find the bookmark matching the one specified when invoking the function. (Bug #36281295)
• Fixed an issue related to audit_log_read performance. (Bug #36281295, Bug #38275868)
• Audit log pruning did not function after removing or renaming a file from the audit log. Now pruning continues in such cases, but with a warning printed in the error log stating that it was not possible to delete the missing audit log file. (Bug #35902913)
• Improved the audit_null example plugin. (Bug #35820314)
• MySQL now calls plugin->deinit() with a valid plugin struct as an argument regardless of the plugin's type. Our thanks to Martin Alderete for the contribution.

Authentication Notes
• Made improvements to the authentication code. (Bug #35325870)
• Improved log messages to provide clear reasons for Access denied errors when using the authentication_ldap_sasl plugin without proxying. (Bug #35317691)
• MySQL LDAP SASL authentication, when used with the GSSAPI method to access an OpenLDAP server, was rejected with the MySQL server error Plugin authentication_ldap_sasl reported: 'LDAP authentication failed or group retrieval failed: LDAP error: Invalid DN syntax', because OpenLDAP did not recognize the root DN used. (Bug #32631511)

C API Notes
• C API applications stalled while receiving results for server side prepared statements.

Character Set Support
• When resolving a call to the REPLACE() function, the character set and collation of the function result are copied from the first argument. The remaining two arguments, if they are literal values, should be converted to this character set, but only the second argument was converted. This fix ensures that the third argument is also converted to the first argument's character set and collation. (Bug #114769, Bug #36562972)
• The internal function my_instr_mb() assumed incorrectly assumption in several places that byte lengths for input strings could be used to short-cut certain decisions. In the case of multibyte character sets and collations, this cannot be done since, under some collation rules, characters with differing byte lengths can be considered equal. In addition, Item_func_locate() used an incorrect byte length. Our thanks to Dirkjan Bussink for the contribution. (Bug #113933, Bug #36277823)
• The internal function Item_func::eq() erroneously treated the two expressions CONVERT(a USING latin1) and CONVERT(a USING utf8mb4) as being equal. (Bug #113506, Bug #116762, Bug #36137690, Bug #37323921)

Compilation Notes
• macOS: CMake no longer tries to use the native ctags on MacOS, and now requires the Homebrew version of it to be installed on the system when building MySQL. (Bug #36590594)
• macOS: Enabled use of gRPC when building MySQL on MacOS. (Bug #36537726)
• macOS: The Xcode version of zlib was removed from the default list of system libraries to use when configuring with -DWITH_SYSTEM_LIBS=ON. (Bug #36537593)
• Microsoft Windows: The BUILD_ALL target did not work when compiling on Windows. (Bug #36424619)
• Microsoft Windows: Excessive RAM usage led to disk swapping when compiling MySQL on Windows using Ninja. (Bug #36399256)
• Interface libraries for librpdserver_shared.so were missing from the bundled protobuf/grpc on Ubuntu 24.04. (Bug #36678790)
• Maintainer mode is now disabled when building the debug version of the server for .deb packages. (Bug #36619757)
• Upgraded the bundled googletest and googlemock sources to version 1.14.0. (Bug #36562482)
• Added a missing dependency on GenError. (Bug #36551721)
• When compiling on Fedora 38, grep -E is now used in place of egrep. (Bug #36507549)
• The version of Boost used for compiling MySQL was upgraded from 1.84.0 to 1.85.0. For more information see the Boost 1.85.0 Release Notes. (Bug #36495694)
• Binaries for Enterprise Linux 8 and 9 are now built using GCC 13. (Bug #36331855)
• Removed linker warnings raised when compiling code that used RapidJSON. (Bug #36322583)
• It is now possible on Linux systems to build MySQL using a bundled tcmalloc library that is provided with the source by specifying -DWITH_TCMALLOC=BUNDLED. This is supported on Linux only. (Bug #36313839)
• The bundled tcmalloc() is now used when building MySQL on Enterprise Linux 8. (Bug #114844, Bug #35674008)
• Removed warnings raised in sql/statement/ed_connection.cc when building on Ubuntu 23.04. (Bug #114436, Bug #36428465)
• Linux aarch64 platform binaries are now built using patchelf --page-size=65536 for compatibility with systems using either 4k or 64k for the page size. (Bug #114233, Bug #36393794)

Component Notes
• When the server was started with an incorrect --basedir option and no --plugin-dir option on UNIX platforms with a manifest and configuration file in place, the keyring component loaded but failed to unload. Now, when loading keyring components from the manifest, if --basedir is specified but --plugin-dir is not specified, we derive the plugin directory from the base directory; if neither of these is specified, we derive the plugin directory from the MySQL installation directory. The same handling also now applies to --datadir. (Bug #36398484)
• The values for component options set using the --loose prefix were not read when the component was installed. (Bug #28341329)

Configuration Notes
• Microsoft Windows: On Windows, MySQL Configurator was updated to support in-place upgrades as per Upgrade Paths. (Bug #36685422)
• Microsoft Windows: On Windows, clicking the [X] close button on a MySQL Configurator wizard's page now yields a confirmation popup if the wizard is busy executing an operation. (Bug #36671317)
• Microsoft Windows: On Windows, MySQL Configurator no longer defines a custom server_type variable in the generated MySQL Server configuration file. This information is now stored in the configurator_settings.xml file. (Bug #36670309)
• Microsoft Windows: On Windows, the Removing Windows Firewall step in MySQL Configurator would fail if the my.ini file was missing a mysqlx_port definition. (Bug #36666260)
• Microsoft Windows: On Windows, if MySQL Configurator failed to find a valid my.ini or my.cnf file from the MySQL Server Installations page, then clicking the Browse button disabled the Next button even when the selected file was valid. (Bug #36395569)
• Microsoft Windows: On Windows, MySQL Configurator now only shows the removal steps if the associated MySQL Server was previously configured. (Bug #36395417)
• Microsoft Windows: On Windows, MySQL Configurator now detects if the existing root user is using the mysql_native_password authentication plugin (removed in MySQL 9.0.0) and prompts to convert root to use the caching_sha2_password authentication plugin before performing a MySQL Server upgrade. (WL #16139)
• Microsoft Windows: For MSI installations on Windows, MySQL Configurator now automatically upgrades MySQL 8.4 LTS installations without user intervention. (WL #16274)

Connection Management Notes
• The conn_delay/Waiting in connection_control plugin stage was not reset after a delay introduced by the connection control plugin which resulted in incorrect monitoring information. (Bug #35205358)

Data Dictionary Notes
• Attempting to upgrade a MyISAM table containing a mix of regular columns and generated columns from MySQL 5.7 to 8.0 or later led to table corruption. (Bug #105301, Bug #33503328)

Data Type Notes
• When a string is converted to a numeric value, any non-numeric data trailing the numeric value should cause an error with strict mode and a warning with any other SQL mode, but in some cases, depending on the length and character set of the string, an invalid string did not raise any of the expected errors or warnings. (Bug #36457756)
• In some cases, casting a double to an integer value used rounding, and in others, with truncation, which led to inconsistent results. Now rounding up is used in all such cases. (Bug #114549, Bug #36481397)
• In some cases, DECIMAL 0 was treated as less than a FLOAT value between 0 and -1. (Bug #114196, Bug #117093, Bug #36361165, Bug #37438582)

Deprecation and Removal Notes
• Replication: MySQL 9.0.0 deprecates transactions which update both transactional tables and nontransactional or noncomposable tables. Such a transaction now causes a deprecation warning to be written to both the client and the error log. Only the InnoDB and BLACKHOLE storage engines are transactional and composable. This means that only the combinations of storage engines shown here do not raise the deprecation warning:
■ InnoDB and BLACKHOLE
■ MyISAM and MERGE
■ performance_schema and any other storage engine
■ TempTable and any other storage engine
Note
NDBCLUSTER is transactional but not composable.
For more information, see Replication and Transactions. (WL #10495)
• The mysql_native_password authentication plugin, deprecated in MySQL 8.0, has been removed, and the server now rejects mysql_native authentication requests from older client programs which do not have CLIENT_PLUGIN_AUTH capability. For backward compatibility, mysql_native_password remains available on the client; the client-side built-in authentication plugin has been converted into a dynamically loadable plugin.
Note
In MySQL 8.0, the default MySQL authentication plugin was changed to caching_sha2_password (see Caching SHA-2 Pluggable Authentication).
These changes also entail the removal of the following server options and variables:
■ The --mysql-native-password server option
■ The --mysql-native-password-proxy-users server option
■ The default_authentication_plugin server system variable
For more information, see Authentication Plugins. (WL #15930)
• The MIN_VALUE and MAX_VALUE columns of the Performance Schema variables_info table are now deprecated, and subject to removal in a future version of MySQL. Instead, you should use the MIN_VALUE and MAX_VALUE columns of the variables_metadata table, which provide the same information. (WL #15585)

Event Scheduler Notes
• Important Change: The following SQL statements relating to events may now be prepared:
■ CREATE EVENT
■ ALTER EVENT
■ DROP EVENT
Positional parameters (? placeholders) are not supported for these statements; you must assemble the text of the statement to be prepared from some combination of string literals, system variables, and user variables. See PREPARE, EXECUTE, and DEALLOCATE PREPARE Statements, and SQL Syntax Permitted in Prepared Statements, for more information. CREATE EVENT Statement, provides a basic example. (Bug #109309, Bug #34875573, WL #16298)

JavaScript Programs
• MySQL 9.0 Enterprise Edition now includes support for stored programs written in JavaScript, such as this simple example created using the CREATE FUNCTION statement and JavaScript code shown here:
CREATE FUNCTION gcd(a INT, b INT) RETURNS INT NO SQL LANGUAGE JAVASCRIPT AS $mle$
let x = Math.abs(a);
let y = Math.abs(b);
while (y) {
var t = y;
y = x % y;
x = t;
}
return x;
$mle$;
• JavaScript Stored Program Creation and Management, describes creation and execution of JavaScript stored programs.
• JavaScript language support includes both stored procedures and stored functions, and is provided by the Multilingual Engine Component (MLE). For more information about determining whether your distribution includes this component, and enabling it, see Multilingual Engine Component (MLE).
• JavaScript language support in MySQL conforms to the ECMAScript 2023 Specification, and uses strict mode by default. Strict mode cannot be disabled. This implementation includes all of the standard ECMAScript library objects such as Object, Function, Math, Date, and String. console.log() and console.error() are also supported.
• Most MySQL data types are supported for JavaScript stored program input and output arguments, as well as for return data types. Strings must use the utf8mb4 character set. MySQL BLOB and TEXT types are supported, as are many MySQL temporal types. JSON is also supported. The VECTOR type is not supported by the MLE component or by JavaScript stored programs. For more information, see JavaScript Stored Program Data Types and Argument Handling, and JavaScript Stored Program Limitations and Restrictions.
• Stored programs written in JavaScript support an SQL and result set API provided by the MLE component. See JavaScript SQL API, and Using the JavaScript SQL API, for more information.
• The MLE component provides a number of session information and management functions including mle_session_state() and mle_session_reset(). You can also view a number of MLE status variables in the output of a statement similar to SHOW STATUS LIKE 'mle%'. See also JavaScript Stored Programs—Session Information and Options.
• For general information about JavaScript stored programs, see JavaScript Stored Programs. (WL #15605, WL #16129, WL #16172, WL #16226, WL #16272, WL #16276)

Keyring Notes
• Made a number of improvements in the keyring code. (Bug #36401550)

Optimizer Notes
• Important Change: ER_SUBQUERY_NO_1_ROW has been removed from the list of errors which are ignored by statements which include the IGNORE keyword. This has been done for the following reasons:
■ Ignoring such errors sometimes led to insertion of NULL into non-nullable columns (for untransformed subqueries), or of no row at all (subqueries using subquery_to_derived).
■ When subqueries were transformed to join with derived tables, the behavior differed from that of untransformed queries.
Following an upgrade to this release, this change can make an UPDATE, DELETE, or INSERT statement which includes the IGNORE keyword raise errors if it contains a SELECT statement with a scalar subquery that produces more than one row.
For more information, see The Effect of IGNORE on Statement Execution. (Bug #110961, Bug #35373406)

Performance Schema Notes
• Group Replication: The following tables did not contain data on replication channels which did not have a configured hostname, such as Group Replication recovery channels:
■ REPLICATION_CONNECTION_STATUS
■ REPLICATION_CONNECTION_CONFIGURATION
■ REPLICATION_APPLIER_CONFIGURATION
■ REPLICATION_APPLIER_STATUS
■ REPLICATION_APPLIER_STATUS_BY_COORDINATOR
■ REPLICATION_APPLIER_STATUS_BY_WORKER
As of this release, these tables contain data for partially configured Group Replication channels. (Bug #36018242)
• Under certain conditions, a race condition could result in the amount of RAM used by TABLE_HANDLES increasing to a maximum of 9GB. (Bug #36170903)
• The PROCESSLIST_INFO column of THREADS was not updated when executing a prepared statement. Thanks to Daniel Lenski and Amazon for the contribution. (Bug #104121, Bug #33057164)
• This release adds two tables to the MySQL Performance Schema, listed here:
■ The variables_metadata table provides general information about system variables. This information includes the name, scope, type, range (where applicable), and description of each system variable recognized by the MySQL server. The MIN_VALUE and MAX_VALUE columns of this table are intended to replace the deprecated MIN_VALUE and MAX_VALUE columns of the variables_info table.
■ The global_variable_attributes table provides information about attribute-value pairs assigned by the server to global system variables.
For more information, see Performance Schema System Variable Tables. (WL #15855)

SQL Syntax Notes
• JSON: You can now save the JSON output from EXPLAIN ANALYZE into a user variable using the syntax shown here:
EXPLAIN ANALYZE FORMAT = JSON INTO @ variable select_stmt
The variable can be used subsequently as a JSON argument to any of MySQL's JSON functions (see JSON Functions). The INTO clause is supported only with FORMAT=JSON, which must be included explicitly. This form of EXPLAIN ANALYZE also supports an optional FOR SCHEMA or FOR DATABASE clause preceding the SELECT statement being analyzed. Statements other than SELECT are not supported.
Note
This feature is available only if the explain_json_format_version server system variable is set to 2; otherwise, attempting to make use of it raises ER_EXPLAIN_ANALYZE_JSON_FORMAT_VERSION_NOT_SUPPORTED.
For more information and examples, see Obtaining Execution Plan Information. (WL #16216)
• MySQL now accepts and enforces inline foreign key specifications (these were previously accepted by the parser, but ignored). MySQL now also accepts implicit references to parent table primary key columns.
Consider the parent table person created by the following statement:
CREATE TABLE person (id SMALLINT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY, name CHAR(60) NOT NULL);
To create a table shirt having a foreign key owner on person, MySQL now accepts and handles correctly any of the CREATE TABLE statements shown here, according to the standard:
CREATE TABLE shirt (id SMALLINT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY, style ENUM ('tee', 'polo', 'dress') NOT NULL, color ENUM ('red', 'blue', 'yellow', 'white', 'black') NOT NULL, owner SMALLINT UNSIGNED NOT NULL, FOREIGN KEY (owner) REFERENCES person (id));
In previous versions of MySQL, only the first of the four statements just shown had the effect of creating a foreign key; the second was parsed, but the REFERENCES clause was ignored. The remaining two statements caused syntax errors.
For more information, see FOREIGN KEY Constraints, as well as FOREIGN KEY Constraint Differences. (Bug #4919, Bug #17943, Bug #102904, Bug #11744902, Bug #11745637, Bug #32613571, WL #16130, WL #16131)

sys Schema Notes
• The performance of the innodb_lock_waits view is improved in this release. (Bug #36337708)

Thread Pool Notes
• Connecting to a thread group that had no connection handler threads stalled. We fix this by making sure that connection handler threads terminate only if there is at least one connection thread left. (Bug #36550125)
• Previous refactoring incorrectly removed the connection locking performed when processing incoming connection requests, which led to a race condition between the thread adding new connections and the connection handler thread processing them. This appeared to cause a situation in which connection requests might be ignored and not processed, so that the connection attempt appeared to hang.
We fix this by taking the connection before processing the queue, and releasing it before waking or creating new threads. (Bug #36548687)
• The Information Schema tables TP_THREAD_GROUP_STATE, TP_THREAD_GROUP_STATS, and TP_THREAD_STATE were deprecated in MySQL 8.0.14. Accessing any of these tables now produces a warning; you should use the equivalent Performance Schema tables instead.
For more information, see Performance Schema Thread Pool Tables. (Bug #36359860)
• It was possible to set the thread_pool_longrun_trx_limit system variable to values outside its stated range. In addition, settings for this variable were not reflected in the output of SHOW VARIABLES or SELECT. (Bug #36347102, Bug #36371145)
• SET PERSIST_ONLY did not work correctly with thread_pool_max_transactions_limit. (Bug #35019884)
• KILL CONNECTION did not work correctly with thread_pool_max_transactions_limit. (Bug #34019954)
• The thread ID was not displayed for client connections in the performance_schema.socket_instances table when using the Thread Pool plugin. (Bug #24796018)

Vector Data Type
• Support is added in this release for a VECTOR column type. A vector is a data structure which consists of a list of entries (4-byte floating-point values) which can be expressed either as a binary string value or a list-formatted string. A VECTOR column is declared with a maximum length or number of entries (in parentheses); the default is 2048, and the maximum is 16383.
• You can create InnoDB tables with VECTOR columns using CREATE TABLE as shown here:
mysql> CREATE TABLE v1 (c1 VECTOR (5000));
Query OK, 0 rows affected (0.03 sec)
• Other storage engines do not support tables with VECTOR columns.
• Vector columns in this release are subject to restrictions, some of which are listed here:
■ A VECTOR column cannot be used as any type of key. This includes primary keys, foreign keys, unique keys, and partitioning keys.
■ Some types of MySQL functions and operators do not accept vectors as arguments. These include but are not limited to numeric functions and operators, temporal functions, full-text search functions, XML functions, bit functions, and JSON functions.
■ Some (but not all) string and encryption functions support VECTOR values. For more complete information, see VECTOR Supported and Unsupported Functions.
■ A VECTOR cannot be compared with any other type, and can be compared with another VECTOR only for equality.
■ MLE JavaScript programs do not support VECTOR columns or values.
■ NDB Cluster does not support VECTOR columns or values.
• Use the VECTOR_DIM() function (also added in MySQL 9.0) to obtain the length of a vector. Functions to convert between representations are available. STRING_TO_VECTOR() (alias TO_VECTOR()) takes a list-formatted representation of a vector and returns the binary string representation; VECTOR_TO_STRING() (alias FROM_VECTOR()) performs the inverse, as shown here:
mysql> SELECT STRING_TO_VECTOR('[2, 3, 5, 7]');
+--------------------------------+
| STRING_TO_VECTOR('[2, 3, 5, 7]') |
+--------------------------------+
| 0x00000040000040400000A0400000E040 |
+--------------------------------+
1 row in set (0.00 sec)
mysql> SELECT VECTOR_TO_STRING(0x00000040000040400000A0400000E040);
+------------------------------------------+
| VECTOR_TO_STRING(0x00000040000040400000A0400000E040) |
+------------------------------------------+
| [2.00000e+00,3.00000e+00,5.00000e+00,7.00000e+00] |
+------------------------------------------+
1 row in set (0.00 sec)
For more information and examples, see The VECTOR Type, and Vector Functions. (WL #16081)

X Plugin Notes
• The system variable caching_sha2_password_digest_rounds could not be set to a non-default value using X Protocol. (Bug #36402455)
• An outdated link to the MySQL documentation in the mysql_function_names unit test source file has been updated. Our thanks to Minha Jeong for the contribution. (Bug #113500, Bug #36137217)

Functionality Added or Changed
• A correlated subquery can now be optimized as an outer left join on a derived table when the subquery contains a LIMIT 1 clause (previously, such a subquery could not be transformed in this way if it contained any LIMIT clause). The LIMIT clause must use a literal 1; if it uses any other value, or a placeholder (?) or variable, the subquery is not eligible for this transformation.
For more information, see Correlated Subqueries. (Bug #36475554, WL #16124)
• Join columns are now included in the output of EXPLAIN FORMAT=JSON. (Bug #36230046)

Bugs Fixed
• Important Change; Replication: When replica_parallel_workers was equal to 1, incorrect logical timestamps for transactions caused the replica to stop with an error, despite the fact that logical timestamps are not relevant when there is only a single applier thread.
Now, the replica logs a warning when replica_parallel_workers = 1, and raises an error only when the value of this system variable is greater than 1. In addition, the format of and information contained in the error message has been improved, and the same message is now used in the output of SHOW REPLICA STATUS as well as in the Performance Schema replication_applier_status table. (Bug #36058442)
• InnoDB: MySQL unexpectedly halted on an UPDATE after an ALTER TABLE operation. (Bug #36571091)
References: This issue is a regression of: Bug #35183686.
• InnoDB: Improved the InnoDB recovery logic to reduce pauses between recovery actions. (Bug #36332645)
• InnoDB: File system operations performed by InnoDB now consistently fsync the parent directory when performing directory altering tasks. (Bug #36174938)
• InnoDB: In debug builds, setting the innodb_interpreter_output debug variable would cause the server to unexpectedly halt. This is now a read-only variable. (Bug #36041032)
• InnoDB: Improved os_innodb_umask handling, and made it read-only. (Bug #35932118)
References: This issue is a regression of: Bug #29472125.
• InnoDB: Removed functionality specific to the Fusion IO atomic write feature, a product that was last available in 2014. (Bug #35072139)
• InnoDB: An InnoDB assertion error referencing an invalid column index was triggered when the column index was valid. (Bug #34800754)
• InnoDB: With an empty XA transaction, shutting the server down after an XA START would cause the server to halt unexpectedly. (Bug #32416819)
• InnoDB: Shutting down the replication applier or binlog applier while processing an empty XA transaction caused the system to unexpectedly halt. (Bug #32416819)
• InnoDB: Removed unnecessary heap usage in the Validate_files::check() function. Our thanks to Huaxiong Song for the contribution. (Bug #115041, Bug #36626203)
• InnoDB: Improved the notify_about_advanced_write_lsn() logic to prevent potential log notification delays. Our thanks to Zongzhi Chen for the contribution. (Bug #114660, Bug #36528707)
• InnoDB: If a partition table was read with innodb_parallel_read_threads=1, read performance greatly decreased from any table after 256 reads. InnoDB behaved as if it reached the maximum capacity of parallel read threads despite not using any. Our thanks to Ke Yu for the contribution. (Bug #114154, Bug #36347408)
• InnoDB: Removed an unnecessary conditional check from get_next_redo_rseg_from_undo_spaces(). Our thanks to Alex Xing for the contribution. (Bug #113640, Bug #36185805)
• InnoDB: The result from a spatial index containing a column with a spatial reference identifier (SRID) attribute was empty. In addition, using FORCE INDEX to force a covering index scan on a spatial index led to an assertion. (Bug #112676, Bug #114200, Bug #35894664, Bug #36361834)
• InnoDB: SELECT ... GROUP BY queries were at least twice as slow with the TempTable engine than the Memory engine. (Bug #107700, Bug #34338001)
• Replication: If a source contained a stored, generated column populated by a JSON function and binlog_row_image was set to MINIMAL, any subsequent update or deletion on the underlying column failed with the following error:
Invalid JSON text in argument 1 to function json_extract: 'The document is empty.'
The replica attempted to re-evaluate the generated column and failed with that error because the underlying column was unavailable. As of this release, stored, generated columns are not re-evaluated when the underlying columns are unavailable. (Bug #36515172)
• Replication: When running GTID-based replication with relay_log_space_limit enabled, a restart of the auto positioning protocol sometimes resulted in an infinite loop, leading to a deadlock in replication. This was because relay_log_space_limit was not honored, not only for transactions whose size exceed this limit, but when the replica could not purge previous logs as well.
To fix this issue, we make the following changes:
■ The receiver respects relay_log_space_limit as set by the user, unless a transaction received by the receiver cannot fit into the purged relay log. Before queuing the received transaction, receiver now checks whether scheduling a full transaction is possible. If not, the receiver performs the following actions:
◦ Sets the flag indicating that receiver is waiting
◦ Rotates the relay log
◦ Waits until it is notified that relay log purge was executed and that the applier has purged all available relay logs; after this, the receiver may queue a transaction without checking the limit again
■ Before moving to the next file, the coordinator checks whether the receiver is waiting for available relay log space. If so, the coordinator forcibly purges the applied logs, including the current relay log file. To purge the current relay log file safely, the coordinator must do the following:
◦ Synchronize all of its workers before moving to the next file
◦ Forcibly update group positions, which is necessary to allow current purging of the relay log
◦ Update the variable read by the receiver which contains relay log filename to which applier was moved
These operations are allowed because we know that receiver waits at a transaction boundary and rotates the relay log before waiting.
(Bug #36507020)
• Replication: Worker jobs now contain information about the relay log file which initiated the transaction, instead of using the default defined by relay_log. (Bug #36395631)
• Replication: The column number returned in the error ER_SERVER_REPLICA_CONVERSION_FAILED was incorrect. It was one less than the actual value. (Bug #36246205)
• Replication: Handling an incident while transactions were being committed to the binary log caused MySQL to wait indefinitely. (Bug #35671897)
• Replication: When using row-based replication with binlog_rows_query_log_events enabled, SQL statements are logged in a Rows_query_log_event which is written before the Table_map_event. Since SQL may contain any binary data including embedded nulls (\0), when using strlen(), snprintf(), and other similar C functions which rely on processing up to the null byte, it was possible for the query string to be truncated at the first null byte, which resulted in processing incomplete data.
We fix this by specifying the length of each Rows_query_log_event such that it is no longer necessary to rely on C-style string methods for length calculations; in addition, while processing the query, we now use functions which do not rely on null termination. We also mark the first byte of a Rows_query_log_event data body as unused. (Bug #109401, Bug #35336260)
• Group Replication: Removed a memory leak from /xcom/gcs_xcom_networking.cc. (Bug #36532199)
• Group Replication: Under certain circumstances, after successfully setting a new primary, group_replication_set_as_primary() waited indefinitely for the operation to complete.
As of this release, a periodic check is performed to ensure the function does not wait unnecessarily. (Bug #36348650)
• Group Replication: Under certain circumstances, if a primary's host experienced network inactivity of 20 seconds or more, the secondaries could stop unexpectedly. (Bug #36306144)
• Group Replication: The MEMBER_ID, MEMBER_HOST, and MEMBER_PORT columns of the REPLICATION_GROUP_MEMBERS table were not always populated for offline members. (Bug #36290046)
• Group Replication: Under certain circumstances, if garbage collection occurred just before a relay log rotation, it could cause the applier to stop applying new transactions on the secondary members.
This was caused by garbage collection incrementing the relay log's last_committed and sequence_number, creating a gap in the recorded sequence_number after the log rotation. The applier was unaffected if the gap occurred anywhere else in the relay log.
As of this release, only last_committed is updated during garbage collection. (Bug #36280130, Bug #36446250)
• JSON: Added missing checks for error handling to NULLIF(), COALESCE(), and the shift (>>) operator. (Bug #113668, Bug #35513196, Bug #36198403)
References: See also: Bug #31358416.
• MySQL NDB ClusterJ: Running the ClusterJ test suite resulted in an error message saying a number of threads did not exist. That was due to some wrong handling of threads and connections, which was corrected by this patch. (Bug #36086735)
• Added a missing error check needed when evaluating the <=> operator. (Bug #36570474)
• Added a missing error check needed for evaluating CASE operators. (Bug #36570439)
• Averages of certain numbers were not always computed correctly. (Bug #36563773)
• Removed redundant assignments to Item::m_table_ref in find_field_in_tables() which led to invalid GROUP BY results and other errors. (Bug #36556725, Bug #36557029)
References: This issue is a regression of: Bug #36533080.
• The following files in strings contained incorrect license information:
■ mb_wc.h
■ ctype-uca.cc
■ ctype-ucs2.cc
■ ctype-utf8.cc
■ dtoa.cc
■ strxmov.cc
■ strxnmov.cc
(Bug #36506181)
• In certain unusual cases, the UpdateXML() function did not process all of its arguments correctly. (Bug #36479091)
• With subquery_to_derived=ON, an outer reference was not replaced in some cases following transformation to a derived table. (Bug #36475633)
References: See also: Bug #36314993.
• A missing check for errors relating to TIME values sometimes led to an assert in sql/item.cc. (Bug #36421511)
• Explaining a query which used FORCE INDEX on a spatial index containing a column with SRID attributes led to an unplanned exit. (Bug #36418426)
• The InnoDB OpenTelemetry metrics (mysql.inno) were not automatically updated. (Bug #36399090)
• This fix addresses two issues:
■ An item that was not yet fixed when searching for an item placed in the GROUP BY list led to an assert in include/sql_string.h.
■ The TIME_FORMAT() function did not handle NULL arguments correctly in all cases.
(Bug #36367313, Bug #36367776)
• Updated BuildRequire rules to align with versions now required for CMake and Bison. (Bug #36343254)
• Removed an unused argument from the internal function MY_COLLATION_HANDLER::strstr(). (Bug #36342997)
• An IN predicate containing EXCEPT ALL set operations yielded the wrong result. (Bug #36332697)
• With optimizer_switch set to subquery_to_derived=on, some queries with ROLLUP were not handled properly. (Bug #36314993)
• A query using WHERE primary_key IN(SELECT constant1 EXCEPT SELECT constant2) returned a differing number of rows depending on the presence or absence of an ORDER BY clause. (Bug #36307622)
• When incrementing the reference count for an expression, underlying expressions within this expression are not looked at. While removing an expression, after decrementing the reference count, even the underlying expressions were examined, which led to unintentional deletion of the underlying expressions. This issue manifested in Item_ref::real_item() as well as in an assert in sql/item.h. We fix this by not looking at the underlying expression unless the current expression contains the only remaining reference. (Bug #36204344, Bug #36356279)
• Under certain conditions, EXPLAIN FORMAT=JSON FOR CONNECTION sometimes led to an unplanned exit. (Bug #36189820)
• In transforming subqueries to derived tables, replacement of a subquery in a HAVING condition failed to use an item reference, which led to an assert in production builds and an unplanned exit in debug builds. This occurred because the optimizer did not correctly detect that the subquery was part of a HAVING condition. (Bug #36079456)
References: This issue is a regression of: Bug #35060385.
• It was possible for MaterializeIterator::load_HF_row_into_hash_map() to exhaust resources while re-reading rows. (Bug #36075756)
• Some CREATE USER statements were not handled correctly. (Bug #36022885)
• In certain cases, a lateral join was not handled correctly. (Bug #35945239)
References: See also: Bug #107700, Bug #34338001. This issue is a regression of: Bug #32644631.
• For a SELECT with ORDER BY and LIMIT, the optimizer first chose a full table scan with a very expensive cost, then performed another check and used the perform_order_index type of path, but this was not reflected by the cost in the optimizer plan. (Bug #35930969)
• Client connections were not alway terminated correctly during shutdown. (Bug #35854919)
• Executing mysqldump on a replica would insert the FLUSH TABLES operation, an operation that writes to the binary log. Now FLUSH LOCAL TABLES is inserted instead to prevent GTID related issues during replication due to these binary log changes.
The workaround was to set the --source-data option to 1 or 2. (Bug #35665076)
References: This issue is a regression of: Bug #33630199.
• All internal ACL bitmask variables are now explicitly 32 bits (uint32_t). (Bug #35507223)
• Events created within stored programs were not always handled correctly. (Bug #35395333, Bug #36402968, Bug #37918920)
References: This issue is a regression of: Bug #17809, Bug #11745618.
• It was not possible to add a functional index on FIND_IN_SET(). (Bug #35352161)
• Removed a memory leak observed while running authentication_kerberos under Valgrind. (Bug #34482788, Bug #36570929)
• The gen_range() function as implemented by the (deprecated) data masking plugin did not always return the correct result.
This issue affected the data masking plugin only, and did not affect the data masking component which supersedes it. (Bug #34163992)
• In some cases, a SELECT constant from an empty table with ORDER BY COUNT(*), when used in a view, did not return any rows. (Bug #115035, Bug #36625752)
• In some circumstances, such when DDL operations were performed on a very large number of tables, the error log was flooded with warnings from background histogram updates; the offending warning was concerning a failure to acquire metadata locks on a table.
To remedy this problem we now throttle messages written to the error log from background histogram update operations, the rate being capped at one message per minute, which should suffice for the user to identify potential problems with background histogram updates. In addition, we downgrade all error events that occur during background histogram updates from errors to warnings. (Bug #114845, Bug #36574298)
• Fixed an erroneous comment in include/my_command.h.
Our thanks to Sho Nakazono for the contribution. (Bug #114507, Bug #36455468)
• The range of error numbers for new errors in MySQL 9 has been designated to begin with 6400. (Bug #114414, Bug #36421351)
• When the character set for arguments to a UDF was specified using component services and the argument values passed did not convert cleanly to the desired character set, the UDF ceased executing and returned SQL NULL. (Bug #114409, Bug #36420251)
• It was possible for a deterministic stored function to return an incorrect result when the function used JOIN ON inside the return statement. If the query needed to be reprepared due to a table metadata caused by, for example, FLUSH TABLES between two executions, the ON clause was sometimes lost. (Bug #114235, Bug #36379879)
• The server rejected a query containing a subquery which referred to a column of the parent table. (Bug #113887, Bug #36262779)
• A query such as the following:
SELECT (SELECT COUNT(1) AS cnt FROM t2 WHERE t2.a = t1.a HAVING cnt > 0) FROM t1;
was transformed to this:
SELECT COALESCE(derived_1_2.cnt, 0) AS cnt FROM t1 LEFT JOIN (SELECT COUNT(1) AS cnt, t2.a AS a from t2 GROUP BY t2.a HAVING (cnt > 0)) AS derived_1_2 ON derived_1_2.a = t1.a;
The presence of a false HAVING condition in the subquery should semantically change the correct result of the scalar subquery from zero to NULL, which happened as expected for the original query, but not for the transformed case. (Bug #113319, Bug #36070647)
• SUM(SUBSTRING()) returned a warning as expected, but SUM(DISTINCT SUBSTRING()) did not. (Bug #113171, Bug #36035064)
• Added the missing mysql-community-libs-compat package for the EL8 and EL9 platforms. (Bug #112949, Bug #35975348)
• SHOW PARSE_TREE CREATE SCHEMA caused a server exit in debug builds.
Note
The SHOW PARSE_TREE statement is available in debug builds only.
(Bug #112883, Bug #35964157)
• A different result was obtained when a column reference argument to CHAR() function was replaced with a CASE() expression that was essentially the same as the column reference. This took place when the CHAR() function was placed in the WHERE clause of an outer join, and the column reference was from the inner table of the outer join. An example of such a query is shown here:
SELECT 1 AS c_0 FROM t0 LEFT JOIN t1 ON t0.vkey = t1.vkey WHERE CHAR(CASE WHEN FALSE THEN t1.vkey ELSE t1.vkey END) NOT LIKE 'X';
A wrong value was obtained when the column reference was used directly; the CHAR() function in the WHERE clause was used to convert the outer join to an inner join, although this is correct only when NULL as the argument implies a NULL result, which is not true of CHAR(). The implementation detail that enforces this conversion is that the function's not_null_tables() property function returns the map bit of the table, but when the column reference was replaced with the CASE() expression, the CASE() did not propagate the not_null_tables() value to keep the outer join from being optimized improperly to an inner join.
We fix this problem by setting the CHAR() function's null_on_null property to false instead of true, which ensures that not_null_tables() returns 0 rather than the table's map bit, so that that the outer join is not converted to an inner join when it should not. (Bug #112397, Bug #36118590)

Original source