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.
[2.0.0] - 2026-06-16
🦀 Breaking Changes
Warning
2.0 removes the suggested ClusterId functionality (PR #290).
Suggested ClusterId Functionality
Prior to this change, the wrapper would generate a unique cluster id based on the connection string and the cluster topology; however, in some cases (such as custom endpoints, IP addresses, and CNAME aliases), the wrapper would generate an incorrect identifier. This change was needed to prevent applications with several clusters from accidentally relying on incorrect topology during failover, which could result in the wrapper failing to complete failover successfully. The ClusterId parameter now defaults to 1, so connections that do not specify a ClusterId share a single topology cache.
Migration
| Number of Database Clusters in Use | Requires Changes | Action Items |
|---|---|---|
| Single database cluster | No | No changes required. |
| Multiple database clusters | Yes | Review all connection strings and add the mandatory ClusterId parameter. See the ClusterId documentation for configuration guidance. |
Warning
This breaking change only impacts customers implementing their own custom plugins. Otherwise no changes are required. 2.0 removes aliases from HostSpec and related APIs (PR #302).
Alias Removal
- Removed aliases from
HostSpec. - Removed
IPluginService.FillAliasesAsync(); useIPluginService.IdentifyConnectionAsync()instead. - Changed
IPluginService.SetAvailability()signature to accept aHostSpecinstead of a collection of host aliases. - Added
IPluginService.RoutedHostSpecto record connection routing events and use it for the current connection info.
If you have custom plugins that use FillAliasesAsync() or rely on HostSpec aliases, update them to use IdentifyConnectionAsync() and the new RoutedHostSpec property.
🪄 Added
- Support for Amazon Aurora Global Databases, including multi-region topology awareness and global writer cluster endpoint recognition (PR #290).
- The Global Database Failover Plugin (
gdbFailover), introducing home-region awareness and configurable failover logic for in-home and out-of-home scenarios (PR #292). - The Global Database Read/Write Splitting Plugin (
gdbReadWriteSplitting), introducing home-region restrictions and optional global write forwarding for read/write splitting connections (PR #293). - IAM and federated (ADFS/Okta) authentication support for Aurora Global Database endpoints, resolving the region via the
DescribeGlobalClustersRDS API (PR #297). - Telemetry for the Global Database and Blue/Green plugins (PR #303).
- Automatically trigger writer failover when connections are incorrectly established as read-only connections (PR #306).
- Documentation:
- ClusterId, Aurora Global Databases, and the GDB plugins (PR #300).
- Database Dialects and Target Connection Dialects (PR #304).