Aurora MySQL database engine updates 2025-09-30 (version 3.10.1, compatible with MySQL 8.0.42)

Fixed an issue with the page latching order sent to reader instances. It can cause the reader instance to restart due to a deadlatch when running SELECT queries over tables being updated by its writer.

Fixed an issue that could cause a reader instance's restart to fail when the writer is running a large number of DDL operations.

Fixed an issue which could cause Aurora MySQL Serverless v2 instances to restart when the innodb_purge_threads parameter was manually configured to a value different from the default. The innodb_purge_threads parameter is now automatically managed for Aurora Serverless v2 instances and cannot be modified.

Fixed an issue with Aurora Serverless V2 scaling that resulted in DB instance restarts by preventing critical memory pages from being swapped out.

Fixed an issue where Aurora out of memory (OOM) avoidance wasn't persisting the configured aurora_oom_response DB parameter value after database restart.

Fixed a race condition which could cause availability issue after a failover in Multi-AZ clusters.

Fixed an issue in logical recovery with aurora_enhanced_binlog enabled that could prevent database restarts in case of aborted transactions.

Resolved an issue where database instances could become unresponsive during Parallel Query (PQ) workloads.

Fixed an issue which prevented users with CONNECTION_ADMIN or SUPER privileges from performing an additional connection beyond max_connections limit, as supported in MySQL Community Edition.

Fixed an issue that can cause continuous database server restarts when the volume was allowed to grow to its maximum size due to certain types of queries being mistakenly allowed after the volume was greater than a certain threshold.

Fixed an issue that can cause the writer instance to become irresponsive if reader instances restart while using Global Write Forwarding or Local Write Forwarding.

Fixed an issue that could cause the writer instance to restart when performing ALTER TABLE in parallel with read queries.

Fixed an issue to improve availability with large volume (> 64 TB) bootstrapping during the parallel export operation.

Fixed an issue that could cause a database instance restart operation to fail if max_user_connections is set to a low value.

Fixed an issue causing inaccurate AbortedClients metrics when multiple connections terminate unexpectedly.

Resolved a race condition that could cause incorrect page reads from the buffer pool during Aurora Serverless scale-down operations or during page eviction from the buffer pool. Ref community Bug#116305.

Fixed an issue that causes unexpected "Internal write forwarding error" on reader instances when write forwarding is enabled.

Fixed an issue which could cause the Previous_gtids binlog event to exclude certain GTIDs with Enhanced Binlog enabled and gtid_mode set to ON or ON_PERMISSIVE.

Fixed an issue which can cause memory management issue when parallel query operations on the table with the blob fields.

Fixed issue where a column with partial JSON updates will fail Parallel Export causing internal fallback to a much slower RDS Export.

Fixed an issue that can cause unexpected instance restarts when there is high concurrent write forwarding workload.

Fixed an issue where Zero Downtime Patching (ZDP) / Zero Downtime Restart (ZDR) could result in DB instance restart while restoring warnings with invalid error codes.

Fixed an issue where temporary binary log files were not being properly cleaned up after transaction rollbacks when using binary logging. This fix prevents unnecessary storage consumption from retained temporary files and in certain cases could also prevent anomalies in the binary log files.

A query of the form SELECT 1 FROM t WHERE CAST(a AS UNSIGNED INTEGER) = 1 AND a = (SELECT 1 FROM t) led to an assertion in item_func.cc. (Bug #36128964)