Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[4.1.0] - 2026-06-17

Added accessible AWS regions support (gdbAccessibleRegions) and configurable monitoring connection priority (monitoringConnectionPriority / gdbMonitoringConnectionPriority) for Global Aurora Databases, allowing deployments with restricted cross-region network reachability to constrain failover, read/write splitting, and topology monitoring to reachable regions (PR #1940).
Added LowestLoadHostSelector and HighestLoadHostSelector host selection strategies that choose hosts based on CPU and replica lag metrics with configurable weights (PR #1969).
Added waitForInitialTopologyMs property to the Aurora Initial Connection Strategy Plugin to distribute concurrent initial reader connections (e.g. connection-pool prefill) across readers instead of routing them all to a single reader (Issue #1968, PR #1977).
Added a self-contained AI configuration assistant skill to help users configure the wrapper with their own AI tools (PR #1941).

Fixed ConfigurationProfile not being propagated to PartialPluginService, which caused internal monitoring workers to lose access to profile-configured plugins (e.g. IAM) and fail permanently after the IAM token TTL expired (Issue #1800, PR #1919).
Fixed target-driver-specific session state (e.g. PostgreSQL currentSchema and readOnly) being lost across connection switches triggered by plugins such as initialConnection, failover2, and read/write splitting (Issue #1787, PR #1921).
Fixed the Blue/Green plugin remaining stuck in IN_PROGRESS after RDS completed switchover, which could exhaust connection pools and require an application restart (Issue #1920, PR #1923).
Fixed the Read/Write Splitting Plugin getting stuck retrying an unreachable reader instead of falling back to the writer (Issue #1324, PR #1966).
Lowered the log level emitted on a successful failover (FailoverSuccessSQLException) from SEVERE to INFO to avoid spurious error alerts in monitoring systems (Issue #1951, PR #1952).
Fixed AbstractMonitor.stop() not shutting down its executor, causing monitor shutdown to block for the full termination timeout (PR #1967).

Deprecated HighestWeightHostSelector in favor of the new load-based host selectors (PR #1969).
The failover plugin now checks whether the current connection requires strict-writer failover mode (PR #1976).
Upgraded jsqlparser to 4.9 (PR #1949).
Dependency updates, including PostgreSQL JDBC driver 42.7.11 (addresses CVE-2026-42198), MySQL Connector/J 9.7.0, MariaDB Connector/J 3.5.9, OpenTelemetry, Jackson, and various AWS SDK modules (PR #1912, PR #1960, PR #1974, PR #1908, PR #1910).
Cleaned up compiler warnings and fixed a failing javadoc check (PR #1934, PR #1964).
CI and test improvements: parallelized standard integration tests, added Gradle caching, sped up the unit test and coverage job, fixed required-check gating for docs-only PRs, improved test resource cleanup, and fixed several flaky integration tests (PR #1915, PR #1916, PR #1918, PR #1931, PR #1933, PR #1935, PR #1963, PR #1967, PR #1970).
Documentation: added a list of other AWS Advanced Wrapper drivers and fixed a README typo and a broken Failover Plugin link (PR #1962, PR #1953).