☁️ The MinIO Plugin

Overview

The minio plugin provides robust and native integration for interacting with S3-compatible object storage systems, such as MinIO or Amazon S3. It allows developers to treat cloud object storage buckets as a first-class resource directly within the microCI pipeline. By integrating reliable mechanisms for uploading build artifacts, fetching necessary dependencies, and managing state data, this plugin ensures that all components are stored in a standardized, versioned, and highly available manner before deployment.

Features

The core capability of this plugin is providing seamless connectivity to S3-compatible storage:

🌐 S3-Compatible Object Storage System Integration

Description: This feature enables the CI pipeline to execute standard object storage operations—including listing contents (ls), uploading files/directories (cp), and downloading objects—using the industry-standard mc (MinIO Client) command structure. The plugin encapsulates these complex CLI calls, making them easily declarative in YAML. Use Case: Essential for cloud-native deployments. Instead of relying on local file systems, pipelines can reliably persist all build outputs (logs, models, configuration files, binaries) into a dedicated storage bucket, ensuring data integrity and traceability regardless of the execution environment.

Setup & Configuration

⚙️ Prerequisites (Authentication)

Before using the minio plugin, ensure that your CI runner has access to valid cloud credentials. Best practice is to configure these as secure environment variables within the microCI platform:

• MINIO_ENDPOINT: The endpoint URL of the MinIO server (e.g., http://storage.mycompany.com:9000).
• MINIO_ACCESSKEY: Your primary access key.
• MINIO_SECRETKEY: Your secret key corresponding to the access key.

📄 Configuration Parameters

Configuration is managed primarily through a shell script block (bash) within the plugin definition, leveraging the powerful and flexible MinIO CLI commands (mc).

Parameter	Type	Description	Required?	Example Value
name	String	The unique name of the plugin (must be minio).	Yes	minio
bash	Multi-line String	Allows running arbitrary MinIO Client (mc) commands. This is the primary method for defining complex object storage operations like copy or list.	No	mc ls mybucket/path/
run_as	String	Specifies the user context under which the plugin step will execute (e.g., user).	No	user
envs	Map	Defines additional environment variables to be available during the execution of the underlying command (e.g., passing bucket credentials).	No	{"MINIO_ACCESSKEY": "..."}

Example Configuration (.microCI.yml)

To demonstrate listing objects and uploading artifacts:

steps:
    - name: "Prepare Artifacts"
      script: make build # Step that generates artifact files (Makefile, etc.)

    - name: "Upload Build Artifacts to MinIO Bucket"
      plugin:
        name: minio
        bash: |
          # 1. List objects in a target bucket for verification
          mc ls myci-artifacts/build-reports/
          # 2. Upload the necessary artifacts from workspace (Makefile)
          mc cp Makefile myci-artifacts/$BUILD_ID/
          # 3. Example: Sync an entire directory recursively
          mc sync ./docs $ARTIFACTORY_BUCKET/documentation/

    - name: "Fetch Configuration From MinIO"
      plugin:
        name: minio
        bash: |
          # Download a required config file from the bucket to local workspace
          mc cp myci-artifacts/$BUILD_ID/config.json .

Examples

This section provides an example of using the minio plugin in a production workflow, demonstrating how to list files, copy data to the bucket, and fetch data back into the workspace for subsequent steps.

steps:
  - name: "Save artifact in bucket"
    description: "Description of this step"
    run_as: user
    envs:
      HOME: /tmp/
    plugin:
      name: minio
      bash: |
        # 1. List files in the bucket for verification
        mc ls microci/bucket_name
        # 2. Copy local artifacts from workspace to the remote bucket
        mc cp Makefile microci/bucket_name/
        # 3. Fetch an artifact from the bucket back into the current working directory (.)
        mc cp microci/bucket_name/image.png .