Aurora MySQL database engine updates 2023-07-25 (version 2.12.0, compatible with MySQL 5.7.40)

Default SSL ciphers used by Aurora MySQL have been updated to exclude the less secure DES-CBC3-SHA values from the SSL_CIPHER database parameter. If you encounter SSL connection issues due to the removal of the DES-CBC3-SHA cipher, please use an applicable secure cipher from the following list, Configuring cipher suites for connections to Aurora MySQL DB clusters. More information on MySQL client Connection Cipher Configuration can be found in the MySQL documentation.

CVE-2023-21963

CVE-2023-21912

CVE-2023-21840

CVE-2023-0215

CVE-2022-43551

CVE-2022-37434

CVE-2022-32221

CVE-2021-36222

CVE-2021-22926

CVE-2021-2169

Fixed an issue in the database activity streams event encryption which may cause database restarts

Fixed two issues which can cause a database restart to fail if it occurred while executing a Data Definition Language (DDL) query

Fixed an issue where connection surges can cause increased query latency or a database instance restart

Fixed an issue which, in rare cases, can cause an Aurora replica to restart during simultaneous execution of large update operations or Data Definition Language (DDL) workloads on the writer instance and read operations on the same set of tables on the Aurora replica

Fixed an issue where connection surges could cause the connection establishment process to take longer to complete or to fail with timeout errors

Fixed an issue where the Advanced Auditing log rotation may reduce the freeable memory, which could lead to the database instance restarting

Fixed an issue which can cause an Aurora MySQL reader instance to restart while executing a query which utilizes an Aurora parallel query execution plan

Fixed an issue that can cause the writer instance to restart while executing the OPTIMIZE TABLE query on a table with a Full Text Search (FTS) index

Fixed an issue which can cause the writer instance in an Aurora global database primary Amazon Region to restart when a SELECT FOR UPDATE query is executed using global write forwarding from an Aurora global database secondary Region

Fixed an issue which can cause an Aurora global database secondary Amazon Region reader instance using global write forwarding to restart when a forwarded implicit commit statement encounters an error

Fast insert isn't enabled in this Aurora MySQL version, due to an issue that can cause inconsistencies when running queries such as INSERT INTO, SELECT, and FROM. For more information on the fast insert optimization, see Amazon Aurora MySQL performance enhancements.

Introduced file management performance optimizations on binlog replicas to help reduce contention when writing to relay log files

Fixed an issue that can cause the buffer_pool_read_requests counter to be reported incorrectly in the information_schema metrics

Fixed an issue that can cause the local storage to fill up when performing LOAD FROM S3 or SELECT INTO S3 operations. The issue can also lead to higher CPU utilization, database restarts due to low memory, and increased latency for these queries.

Fixed an issue where DB instances using binary log replication may experience an increase in CPU utilization and connection failures when multiple binary log replication consumers are attached

Fixed an issue where the SSL server status variables weren't being populated

Fixed an issue where the Data Manipulation Language (DML) statements executing duplicate writes could lead to excessive error logging and increased query latencies

Upgraded the time zone definitions to the IANA 2023c version

Added support for enabling and disabling session-level binary logging. See Stored Procedures - Replicating in the Amazon Aurora User Guide

Added support for setting the session-level binary log format. See Stored Procedures - Replicating in the Amazon Aurora User Guide

Fixed an issue where setting the aurora_disable_hash_join parameter to 1 or ON might not prevent the optimizer from using a hash join

Fixed an issue involving index scans where an inaccurate result might be returned when executing a SELECT query with the GROUP BY clause and the aurora_parallel_query parameter turned ON

Fixed an issue which, in rare cases, can cause an Amazon Aurora reader instance to restart when accessing a table which has large update or Data Definition Language (DDL) operations running concurrently on the writer instance

Fixed an issue that can cause the buffer_pool_read_requests counter to be reported incorrectly in the information_schema metrics

Fixed an issue that can cause a binlog replica to restart if the system variable server uuid of the source is missing or has an invalid value

Fixed an issue to prevent InnoDB statistics from getting stale, which can sometimes generate a sub-optimal query execution plan that may lead to an increase in the query execution time

Fixed an issue wherein the AuroraGlobalDBRPOLag CloudWatch metrics always displayed zero regardless of the user workload

To perform a minor version upgrade for an Aurora global database from Aurora MySQL version 2.07 or 2.11 to Aurora MySQL version 2.12 or higher, refer to Upgrading Aurora MySQL by modifying the engine version.

Fixed an issue which can cause higher CPU utilization due to background TLS certificate rotation (Community Bug Fix #34284186)

Scan batching.

The CREATE TABLESPACE SQL statement

Group replication plugin

Increased page size

InnoDB buffer pool loading at startup

InnoDB full-text parser plugin

Multisource replication

Online buffer pool resizing

Password validation plugin

Query rewrite plugins

Replication filtering

X Protocol