Aurora MySQL database engine updates 2025-11-13 (version 3.11.0, compatible with MySQL 8.0.43)

CVE-2025-50077

CVE-2025-50078

CVE-2025-50079

CVE-2025-50080

CVE-2025-50082

CVE-2025-50083

CVE-2025-50084

CVE-2025-50085

CVE-2025-50086

CVE-2025-50087

CVE-2025-50088

CVE-2025-50091

CVE-2025-50092

CVE-2025-50093

CVE-2025-50094

CVE-2025-50096

CVE-2025-50097

CVE-2025-50099

CVE-2025-50101

CVE-2025-50102

CVE-2025-53023

CVE-2025-50098

CVE-2025-50100

CVE-2025-50104

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

Fixed an issue that could cause the writer instance to restart when global write forwarding or local write forwarding is being disabled.

Fixed an issue that could cause writer to become unresponsive when write forwarding is being disabled while two or more replicas are forwarding requests.

Fixed an issue that could cause an engine restart when running KILL <query-id> after running EXPLAIN FOR CONNECTION <query-id> on a running parallel query.

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

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

Fixed an issue that 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 can't be modified.

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 that could cause the writer instance to become unresponsive if reader instances restart while using Global Write Forwarding or Local Write Forwarding.

Fixed an issue that could cause engine restart during zero-downtime patching (ZDP)/zero-downtime restart (ZDR), when preserving SSL/TLS connections with active transactions.

Fixed an issue that, under rare conditions, could cause changes on the writer instance to be sent to the reader instance in an order that's incompatible with the query-processing threads. This can lead to a deadlock between query-processing thread and replication-apply thread, which in turn can cause the reader instance to restart.

Fixed an issue to prevent unnecessary database server restarts that were occurring due to the monitoring agent's incorrect health assessment.

Fixed an issue where writes to the database may stall while executing a long running transaction leading to a database restart or a major version upgrade to fail.

Fixed an issue with replicas incorrectly being restarted when joining the writer.

Optimized database performance for database instances with large Aurora Storage volumes.

Fixed an issue in the range optimizer where constant expressions weren't recognized when calculating range bounds. (Community Bug #112737)

Fixed an issue where exporting a DB snapshot to Amazon S3 performs slower when exporting tables containing JSON columns with null values.

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

Improved write IOPS performance when system variable innodb_flush_log_at_trx_commit is set to 0.

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.

Automatically disable aurora_oom_response actions (except print, if configured) when aurora_oom_response fails to resolve memory pressure after a threshold time (in the order of few mins).

Fixed an issue where a column with partial JSON updates fails parallel export causing internal fallback to a much slower Amazon RDS export.

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

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

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 where exporting a DB snapshot to Amazon S3 will be slower when exporting tables containing secondary indices with generated expressions.

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

Added a new global variable aurora_lambda_request_timeout to allow users configure Amazon Lambda request timeouts (default: 10 seconds). For more information on invoking a Lambda function from an Aurora MySQL DB cluster see Invoking a Lambda function from an Aurora MySQL DB cluster.

Fixed an issue where local write forwarding stops working after the database instance restarted with zero-downtime restart.

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

Improved parallel export performance by optimizing the bootstrap process for large volumes (>64 TB), reducing overall export operation time.

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

Fixed an issue that can cause DB cluster exports to take significantly longer than expected when there are tables larger than 14 TB.

Fixed an issue that caused inaccurate tracking of parallel query requests while running EXPLAIN ANALYZE statements where Aurora_pq_request_in_progress counter wasn't updated accurately.

Fixed an issue that, under uncommon conditions, can cause the database instance to restart when the database volume is close to its maximum allowed size.

Fixed an issue that can cause a database instance restart when restoring connections during zero-downtime restart and zero-downtime patching.

Fixed an issue where a query of the form SELECT 1 FROM t WHERE CAST(a AS UNSIGNED INTEGER) = 1 AND a = (SELECT 1 FROM t) leads to an assertion failure in item_func.cc. (Community Bug Fix #36128964)

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)