This commit introduces a new queue system using p-limit, addressing resource issues and improving job cancellation capabilities. Key changes include:
- Removal of Redis dependency, allowing for in-memory queue management.
- Implementation of per-server queues with ordered processing based on server concurrency settings.
- Addition of helper functions for job management and status retrieval, ensuring backward compatibility with existing API endpoints.
- Updates to database schema to support server concurrency settings.
The legacy BullMQ code has been retained for compatibility but is no longer in active use.
- Introduced a new column `railpackVersion` in the application table with a default value of '0.2.2'.
- Updated the application form to include a field for `railpackVersion` when the build type is set to railpack.
- Adjusted the build process to utilize the specified `railpackVersion` dynamically.
- Enhanced validation schema to accommodate the new field.
- Introduced a new boolean field `previewRequireCollaboratorPermissions` in the application schema to enforce permission checks for preview deployments.
- Updated the UI to include a toggle for this setting in the preview deployment settings.
- Enhanced GitHub deployment handler to validate PR authors against the required permissions, blocking unauthorized deployments and providing security notifications.
- Added SQL migration to update the database schema accordingly.
- Introduced a new boolean column "isolatedDeploymentsVolume" to the "compose" table with a default value of false.
- Updated existing records to set "isolatedDeploymentsVolume" to true.
- Modified related functions to handle the new column for improved deployment isolation management.
- Updated the rollbacks schema to include mounts, ports, and optional registry information in the full context.
- Refactored the rollback service to utilize the full context for improved rollback operations, ensuring all necessary configurations are applied.
- Exported the getAuthConfig function for better accessibility in the application context.
- Introduced a new ENUM type "publishModeType" with values 'ingress' and 'host'.
- Added "publishMode" column to the "port" table with a default value of 'host'.
- Updated related metadata files to reflect the changes.
- Introduced a new appName column in the volume_backup table to improve application identification.
- Updated the volume backup schema to ensure appName is a required field.
- Enhanced the deployment service to support volume backup creation, integrating appName into the deployment process.
- Added validation for volume backup creation to ensure proper handling of appName and related data.
- Added support for volume backups in the deployment management interface by introducing a new volumeBackupId field in the deployment schema.
- Updated the volume backup schema to include relationships with deployments, allowing for better management and tracking of volume backups.
- Enhanced API routes to include application data when querying volume backups, improving the data returned for related entities.
- Updated UI components to reflect the new volume backup features and ensure seamless integration with existing functionalities.
- Modified the volume backup schema to enforce non-null constraints on volumeName and added serviceName.
- Removed unnecessary fields (type, hostPath) from the schema and updated related API and form handling.
- Enhanced form validation to ensure required fields are properly checked.
- Updated the UI components to reflect changes in the volume backup management interface.
- Added components for handling and displaying volume backups in the dashboard.
- Created API routes for managing volume backups, including create, update, delete, and list operations.
- Introduced database schema for volume backups, including necessary fields and relationships.
- Updated the application to integrate volume backup features into the service management interface.
- Implemented a new mutation to kill a running deployment process by its PID.
- Updated the deployment schema to include a PID field.
- Enhanced the deployment service to handle process termination and status updates.
- Modified the deployment scripts to echo PID and schedule ID for better tracking.
- Added error handling for the kill process operation.
- Added refined validation for the 'path' field to prevent directory traversal attacks and unauthorized access.
- Implemented checks for null bytes and ensured paths start with the MAIN_TRAEFIK_PATH constant.
strip path functionality to compose
- Add internalPath field to route requests
to different paths internally
- Add stripPath option to remove external
path prefix before forwarding
- Improves validation for stripPath (requires
non-root path) and internalPath (must start
with /)
- Updated fullContext type in rollbacks schema to include Application and Project types.
- Refactored createRollback function to separate fullContext from input and handle it more efficiently.
- Integrated environment variable preparation into the rollback process.
- Introduced a new userId column in the git_provider table to associate git providers with users.
- Updated the foreign key reference for userId to point to the users_temp table instead of the account table.
- Modified the UnauthorizedGitProvider component to include a dialog action for disconnecting repositories, enhancing user experience.
- Added a migration script to update existing git providers with the new userId values based on the organization owner.
- Add internalPath and stripPath fields to domain schema
- Implement UI controls for configuring internal path routing
- Create Traefik middleware for path manipulation (addPrefix/stripPrefix)
- Support different external and internal paths for applications
- Enable path stripping for cleaner URL forwarding
This allows applications to be accessed via external paths while maintaining
different internal routing structures, useful for microservices and legacy
applications that expect specific path prefixes.
- Added SQL script to set default value for "logCleanupCron" in "user_temp" table.
- Updated existing records with NULL "logCleanupCron" to the new default value.
- Updated user schema to reflect the default value for "logCleanupCron".
- Enhanced log cleanup functionality with error handling and logging.
- Eliminated the "limitRollback" property from the application schema and the ShowRollbackSettings component, streamlining rollback configuration.
- Updated the database schema to drop the "limitRollback" column from the "application" table, ensuring consistency across the application.
- Introduced a new "fullContext" JSONB column in the "rollback" table to store additional context for rollbacks.
- Removed the "env" column from the "rollback" table to streamline data management.
- Updated the rollbacks service to handle the new "fullContext" field during rollback creation.
- Adjusted the application service to eliminate environment variable handling in rollback operations.
- Updated Tailwind configuration for responsive design.
- Modified the ShowDeployments component to include rollback settings and actions.
- Introduced a new "rollback" table in the database schema with foreign key relationships.
- Updated deployment and application schemas to support rollback features.
- Added rollback mutation to the API for initiating rollbacks.
