Registry

Registry is the core component of Dewy that handles application and file version management. Dewy automatically detects the latest version based on semantic versioning and achieves continuous deployment.

Supported Registries

Dewy supports the following registry types:

GitHub Releases (ghr://): GitHub release functionality
AWS S3 (s3://): Amazon S3 storage
Google Cloud Storage (gs://): Google Cloud storage
gRPC (grpc://): Custom gRPC server

Common Options

Common options available for all registries:

Option	Type	Description
`pre-release`	bool	Whether to include pre-release versions
`artifact`	string	Explicitly specify artifact names that are not automatically selected

GitHub Releases

The most common method for using GitHub releases as a registry.

Basic Configuration

# Basic format
ghr://<owner>/<repo>

# Example
dewy server --registry ghr://linyows/myapp -- /opt/myapp/current/myapp

Environment Variables

export GITHUB_TOKEN=ghp_xxxxxxxxxxxxxxxxxxxx

Set GitHub Personal Access Token or GITHUB_TOKEN for GitHub Actions.

Examples with Options

# Include pre-release versions (staging environment)
dewy server --registry "ghr://linyows/myapp?pre-release=true"

# Specify specific artifact
dewy server --registry "ghr://linyows/myapp?artifact=myapp-server.tar.gz"

# Use both options
dewy server --registry "ghr://linyows/myapp?pre-release=true&artifact=myapp-server.tar.gz"

Automatic Artifact Selection

When artifact names are not specified, Dewy automatically selects using the following rules:

File names containing current OS (linux, darwin, windows)
File names containing current architecture (amd64, arm64, etc.)
Select the first matching artifact

Example: In Linux amd64 environment, myapp_linux_amd64.tar.gz is automatically selected.

important

For newly created releases, there is a 30-minute grace period considering CI/CD artifact build time. During this period, "artifact not found" errors are not notified.

AWS S3

S3-compatible storage can be used as a registry.

Basic Configuration

# Basic format
s3://<region>/<bucket>/<path-prefix>

# Example
dewy server --registry s3://ap-northeast-1/my-releases/myapp

Environment Variables

export AWS_ACCESS_KEY_ID=AKIAIOSFODNN7EXAMPLE
export AWS_SECRET_ACCESS_KEY=wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY

# Optional: Endpoint URL (for S3-compatible services)
export AWS_ENDPOINT_URL=https://s3.isk01.sakurastorage.jp

Object Path Rules

Objects in S3 must be arranged in the following structure:

<path-prefix>/<semver>/<artifact>

Actual example:

my-releases/myapp/v1.2.4/myapp_linux_amd64.tar.gz
my-releases/myapp/v1.2.4/myapp_linux_arm64.tar.gz
my-releases/myapp/v1.2.4/myapp_darwin_arm64.tar.gz
my-releases/myapp/v1.2.3/myapp_linux_amd64.tar.gz
my-releases/myapp/v1.2.3/myapp_linux_arm64.tar.gz
my-releases/myapp/v1.2.3/myapp_darwin_arm64.tar.gz

Google Cloud Storage

Google Cloud Storage can be used as a registry.

Basic Configuration

# Basic format
gs://<bucket>/<path-prefix>

# Example
dewy server --registry gs://my-releases-bucket/myapp

Environment Variables

# Use service account key
export GOOGLE_APPLICATION_CREDENTIALS=/path/to/service-account.json

Google Cloud SDK credentials or Workload Identity can also be used.

Object Path Rules

Arranged in the same structure as S3:

myapp/v1.2.4/myapp_linux_amd64.tar.gz
myapp/v1.2.4/myapp_darwin_arm64.tar.gz
myapp/v1.2.3/myapp_linux_amd64.tar.gz

gRPC

Custom gRPC servers can be used as registries.

Basic Configuration

# Basic format
grpc://<server-host>

# Example
dewy server --registry grpc://registry.example.com:9000

# Without TLS
dewy server --registry "grpc://localhost:9000?no-tls=true"

Features

gRPC server dynamically provides artifact URLs
pre-release and artifact options cannot be used
Flexible control with custom logic is possible

Semantic Versioning

Dewy performs version management compliant with semantic versioning (semver).

Supported Formats

# Standard versions
v1.2.3
1.2.3

# Pre-release versions
v1.2.3-rc
v1.2.3-beta.2
v1.2.3-alpha.1

Using Pre-release Versions

# Production environment (stable versions only)
dewy server --registry ghr://myorg/myapp

# Staging environment (including pre-release versions)
dewy server --registry "ghr://myorg/myapp?pre-release=true"

CI/CD Pipeline Usage

# GitHub Actions
- name: Deploy with Dewy
  env:
    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
  run: |
    dewy server --registry ghr://${{ github.repository }} \
      --log-level info --port 8080 -- /opt/app/current/app

Multi-stage Deployment

# Staging environment
ENVIRONMENT=staging dewy server \
  --registry "ghr://myorg/myapp?pre-release=true" \
  --notifier "slack://staging-deploy?title=myapp-staging"

# Production environment
ENVIRONMENT=production dewy server \
  --registry "ghr://myorg/myapp" \
  --notifier "slack://prod-deploy?title=myapp-prod"

Troubleshooting

Artifact Not Found

Check version tags: Are they compliant with semantic versioning?
Check artifact names: Do they contain OS/architecture?
Check permissions: Are authentication credentials correctly configured?

Debug Methods

# Enable debug logs to check details
dewy server --registry ghr://owner/repo --log-level debug

Registry is a crucial component at the core of Dewy's operation. Select the appropriate registry type for your use case and build an efficient deployment environment.