Skip to content

CHANGELOG.md

Latest commit

 

History

History
301 lines (197 loc) · 10.3 KB

CHANGELOG.md

File metadata and controls

301 lines (197 loc) · 10.3 KB
Raw

Changelog

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

[Unreleased]

Added

  • Added a repository notice that the @udx/worker-deployment CLI is entering maintenance mode and new workflow work should prefer repo-owned Rabbit/worker contracts and dev.kit guidance.

Changed

  • Disabled the automated npm release workflow to prevent repo-driven package releases while the currently published CLI remains available.

[3.0.4] - 2026-02-05

Fixed

  • Fix YAML key listing to output real newlines so env vars are passed correctly.
  • Quote command display in deploy make target to avoid shell parse errors.
  • Update default deploy config and docs to use command + args.

[3.0.3] - 2026-02-05

Changed

  • Change release branch from master to latest in npm release workflow

[3.0.2] - 2026-02-05

Removed

  • Removed gh_token from npm release workflow

[3.0.1] - 2026-02-05

Fixed

  • Bump version to 3.0.1 to resolve immutable tag issue

[3.0.0] - 2026-02-02

Breaking Changes

  • Replaced worker-config and worker-run binaries with a single worker CLI and subcommands.
  • Commands now use worker <subcommand> (for example, worker config, worker run).

Added

  • New CLI subcommands: config, run, gen, and images.
  • Documentation for config/run/gen/images plus deploy configuration and CLI env docs.
  • Worker generation templates and repo scaffolding assets.
  • CI build and release scripts plus an npm release workflow.

Changed

  • Package entrypoints and release scripts updated for the unified CLI.
  • Test harness reorganized with new command coverage.

[2.5.0] - 2024-12-02

🚀 Added - Network and Container Name Support

Added

  • Network Configuration - Containers can now be deployed with custom Docker networks
    • Useful for connecting containers to existing networks or using host networking
    • If not specified, Docker uses default bridge network
  • Container Name Configuration - Containers can now have custom names
    • Specify custom container names via container_name field
    • If not specified, Docker auto-generates a random name
    • Useful for container management and inter-container communication

Enhanced

  • Comprehensive Config Template - Updated deploy.yml template with:
    • All supported configuration options documented
    • Format specifications for volumes, ports, network, and container names
    • Practical examples for each option
    • Clear comments explaining optional vs required fields

Technical Details

  • Added NETWORK parsing and formatting with --network flag in deploy.sh
  • Added CONTAINER_NAME parsing and formatting with --name flag in deploy.sh
  • Updated deploy.mk to include network and container name in docker run commands
  • Both run and run-it targets display network and container name settings
  • Added test coverage for both network and container_name configurations

Benefits

  • Better Networking - Connect containers to custom networks or use host networking
  • Easier Management - Named containers are easier to identify and manage
  • Complete Docker Support - Now supports all common Docker run options
  • Better Documentation - Config template serves as comprehensive reference

[2.4.0] - 2025-11-04

🎯 Enhanced - Added Ports Support

Changed

  • Ports Parameter now supported - Containers can now can be deployed with specified ports
    • If ports is not specified in deploy.yml, the container's default ports will be used

Technical Details

  • Updated deploy.mk with conditional logic to only include ports in docker run when present
  • Both run and run-it targets handle empty ports gracefully

Benefits

  • More Flexible - Can deploy containers that have their own default ports
  • Better UX - Clear indication when using container defaults vs custom ports

[2.3.0] - 2025-10-31

🎯 Enhanced - Optional Command Parameter

Changed

  • Command Parameter Now Optional - Containers can now run with their default CMD/ENTRYPOINT
    • If command is not specified in deploy.yml, the container's default command will be used
    • Useful for containers with well-defined default behavior
    • Display shows <using container default> when no command is specified
    • Dry-run output correctly reflects whether command is present or using default

Added

  • New Example - Added documentation example showing usage without command parameter
  • Enhanced Guidance - Updated config template and help text to indicate command is optional

Technical Details

  • Removed validation that required command field in deploy.sh
  • Updated deploy.mk with conditional logic to only include command in docker run when present
  • Both run and run-it targets handle empty command gracefully
  • Config template includes helpful comment explaining optional nature

Benefits

  • More Flexible - Can deploy containers that have their own default commands
  • Cleaner Configs - No need to specify command when container default is sufficient
  • Better UX - Clear indication when using container defaults vs custom commands

[2.2.0] - 2025-10-30

🎯 Enhanced - Simplified Authentication Architecture

Changed

  • Unified Authentication Approach - Simplified GCP credential handling by delegating to worker image
    • All credential types (service account keys, workload identity tokens, impersonation) now use single GCP_CREDS environment variable
    • Worker's gcp.sh module handles credential type detection, normalization, and GOOGLE_APPLICATION_CREDENTIALS setup
    • Removed complex dual-variable approach for impersonation
    • Cleaner, more maintainable implementation

Technical Details

  • Simplified src/providers/gcp.mk to use unified GCP_CREDS variable for all authentication methods
  • Worker image now handles all credential normalization (private_key escaping, type detection, etc.)
  • Reduced complexity in deployment tool by centralizing auth logic in worker
  • Maintains full backward compatibility with all three authentication methods

Benefits

  • Simpler codebase - Less authentication logic in deployment tool
  • Single source of truth - Worker image controls authentication behavior
  • Better maintainability - Authentication changes only need to happen in worker image
  • Consistent behavior - All tools using worker get same authentication handling

[2.1.1] - 2025-10-28

Fixed

  • Adjusted environment variable settings and volume mounts for GCP authentication

[2.1.0] - 2025-10-21

🎯 Enhanced - Full Terraform/SDK Support for Impersonation

Changed

  • Service Account Impersonation - Now fully compatible with Terraform, SDKs, gcloud CLI, and all Google Cloud client libraries
    • Uses user's Application Default Credentials (ADC) as source for impersonation
    • Generates proper impersonated_service_account credential file (official Google format)
    • Sets both GOOGLE_APPLICATION_CREDENTIALS (for Terraform/SDKs) and CLOUDSDK_AUTH_ACCESS_TOKEN (for gcloud)
    • Dual authentication ensures all tools work seamlessly
    • Eliminates the need for service account key files even when using Terraform

Added

  • ADC Integration - Automatically uses ~/.config/gcloud/application_default_credentials.json
  • Fallback Support - Falls back to access token if ADC not configured
  • Setup Guidance - Clear instructions for gcloud auth application-default login

Technical Details

  • Creates impersonated_service_account type credential file using user's ADC as source
  • Mounts credential file as read-only in container
  • Passes access token as environment variable for gcloud CLI compatibility
  • Follows official Google Cloud authentication best practices
  • Token remains valid for 1 hour

[2.0.2] - 2025-10-21

Fixed

  • Impersonation Error Handling - Improved error detection and display for gcloud impersonation failures
  • Better filtering of WARNING messages from gcloud output

[2.0.1] - 2025-10-21

Changed

  • README Simplification - Prioritized impersonation method, merged auth sections, focused on worker examples

[2.0.0] - 2025-10-21

🎉 Major Release - Complete Authentication Overhaul

Added

  • Service Account Impersonation - Use your gcloud credentials to impersonate service accounts
    • Automatic token generation on host
    • No key files needed
    • Temporary tokens (1 hour validity)
    • Works with gcloud CLI commands
  • Modular Provider System - Extensible architecture for multiple cloud providers
    • GCP provider with full authentication support
    • Ready for AWS, Azure providers
  • Config-Based Authentication - Specify custom credential paths in deploy.yml
    • service_account.key_path - Custom key file location
    • service_account.token_path - Custom token file location
    • service_account.email - Service account impersonation
  • Enhanced Error Messages - Clear, actionable error messages with fix suggestions
  • Improved Security - Docker commands no longer echo sensitive tokens to terminal

Changed

  • Authentication Priority - Clearer credential detection order
    1. Config-specified paths
    2. Default file paths (gcp-key.json, gcp-credentials.json)
    3. Service account impersonation
  • Simplified CLI - Only 2 commands: worker-config and worker-run
  • Better Documentation - Comprehensive README with all authentication methods

Fixed

  • UID/GID Mismatch - Resolved permission issues with credential files
  • Token Filtering - Properly filters gcloud WARNING messages from tokens
  • Error Handling - Improved error detection and reporting for impersonation

Technical Improvements

  • Modular makefile architecture (src/providers/gcp.mk)
  • Automatic credential detection and validation
  • Secure token handling (no hardcoded OAuth credentials)
  • Read-only credential mounts where appropriate

[1.3.0] - 2025-10-20

Added

  • Dry-run capability with --dry-run flag
  • Docker and yq availability checking
  • Configuration validation for required fields

Improved

  • Enhanced error messages with colors
  • Better volume path resolution
  • Comprehensive test coverage

[1.0.0] - Initial Release

Features

  • YAML-based configuration
  • Docker container execution
  • Basic GCP credential mounting
  • Volume and environment variable management