- Added methods to check for existing Raft state and clear it if necessary, allowing for a clean join to the cluster.
- Enhanced automatic recovery logic to detect stale Raft state and clear it when peers have higher log indexes.
- Improved logging for Raft state operations to provide better visibility during cluster management.
- Introduced a new script `dev-kill-all.sh` to handle graceful shutdown of development processes and cleanup of stale PID files.
- Updated Makefile to include a `kill` command that utilizes the new script for improved process management.
- Enhanced the shutdown process to verify that required ports are free after termination of processes.
- Consolidated development commands into a new `dev` command group for better organization.
- Introduced a `prod` command group to manage production environment operations.
- Updated Makefile to simplify the development environment setup and improve logging.
- Enhanced README to clarify the development process and health check requirements.
- Removed deprecated configuration and service management commands to streamline the CLI interface.
- Added structured logging for RQLite components, including cluster discovery and leadership processes.
- Introduced methods for preparing the data directory and launching the RQLite process, improving code organization.
- Implemented exponential backoff for leadership checks to reduce log noise and improve reliability.
- Enhanced peer health tracking and membership update logic to streamline cluster synchronization and recovery.
- Introduced TriggerPeerExchange method to facilitate manual peer exchange for RQLite metadata.
- Implemented performPreStartClusterDiscovery to ensure coordinated recovery by building peers.json before RQLite startup.
- Added validation for node ID consistency with raft address during RQLite startup.
- Enhanced logging for cluster recovery processes and peer discovery progress.
- Updated cluster synchronization logic to improve reliability during node recovery scenarios.
- Added ClusterDiscoveryService to manage peer discovery and synchronization for RQLite nodes.
- Introduced new configuration options for cluster synchronization intervals, peer inactivity limits, and minimum cluster size.
- Enhanced validation logic to ensure proper configuration of cluster parameters.
- Implemented metrics collection for cluster health and peer status, improving monitoring capabilities.
- Updated RQLiteManager to integrate with the new discovery service, allowing for dynamic leadership and cluster joining logic.
- Added steps to generate node2.yaml, node3.yaml, and gateway.yaml configurations during the full stack initialization process.
- Implemented checks to prevent overwriting existing configuration files unless the --force flag is used.
- Improved output messages to provide clear feedback on generated configurations and bootstrap information.
- Removed old pkg/gatway storage and migrated to new rqlite
- Updated readme
- Created new rqlite implementation
- Updated changelog
- Fixed migration error on migrations.go applySQL
- Created new rqlite folder
- Created rqlite adapter, client, gateway, migrations and rqlite init
- Updated node.go to support new rqlite architecture
- Updated readme
- Updated version
