website-and-documentation/content/en/docs/edp/forgejo/_index.md

title	linkTitle	weight	description
Forgejo	Forgejo	5	Forgejo provides source code management, project management, and CI/CD automation for the EDP.

The internal service is officially designated as the Edge Developer Platform (EDP). It is hosted at edp.buildth.ing. The domain selection followed a democratic team process to establish a unique identity distinct from standard corporate naming conventions.

Technical Architecture & Deployment

Infrastructure Stack

The platform is hosted on the Open Telekom Cloud (OTC). The infrastructure adheres to Infrastructure-as-Code (IaC) principles.

• Deployment Method: The official Forgejo Helm Chart is deployed via ArgoCD.
• Infrastructure Provisioning: Terraform is used to provision all underlying OTC services, including:
  • Container Orchestration: CCE (Cloud Container Engine): Kubernetes
  • Database: RDS (Distributed Cache Service): PostgreSQL
  • Caching: DCS (Distributed Cache Service): Redis
  • Object Storage: OBS (Object Storage Service, S3-compatible): for user data (avatars, attachments).
  • Search: CSS (Cloud Search Service): Elasticsearch

The "Self-Replicating" Pipeline

A key architectural feature is the ability of the platform to maintain itself. A Forgejo Action can trigger the deployment script, which runs Terraform and syncs ArgoCD, effectively allowing "Forgejo to create/update Forgejo."

graph TD
    subgraph "Open Telekom Cloud (OTC)"
        subgraph "Control Plane"
            Dev[DevOps Engineer] -->|Triggers| Pipeline[Deployment Pipeline]
            Pipeline -->|Executes| TF[Terraform]
        end

        subgraph "Provisioned Infrastructure"
            TF -->|Provisions| CCE[(CCE K8s Cluster)]
            TF -->|Provisions| RDS[(RDS PostgreSQL)]
            TF -->|Provisions| Redis[(DCS Redis)]
            TF -->|Provisions| S3[(OBS S3 Bucket)]
            TF -->|Provisions| CSS[(CSS Elasticsearch)]
        end

        subgraph "Application Layer (on CCE K8s)"
            Pipeline -->|Helm Chart| Argo[ArgoCD]
            Argo -->|Deploys| ForgejoApp[Forgejo]
        end

        CCE -- Runs --> Argo
        CCE -- Runs --> ForgejoApp
        ForgejoApp -->|Connects| RDS
        ForgejoApp -->|Connects| Redis
        ForgejoApp -->|Connects| S3
        ForgejoApp -->|Connects| CSS
    end

Migration History

The initial environment was a manual setup on the Open Sovereign Cloud (OSC). Once the automation stack (Terraform/ArgoCD) was matured, the platform was migrated to the current OTC environment.

Application Extensions

Core Functionality

Beyond standard Git versioning, the platform utilizes:

• Releases: Hosting binaries for software distribution (e.g., Edge Connect CLI).
• CI/CD: Extensive pipeline usage for build, test, and deployment automation.
• Note on Issues: While initially used, issue tracking was migrated to JIRA to align with the broader IPCEI program standards.

GARM (Git-based Actions Runner Manager)

The primary technical innovation was the integration of GARM to enable ephemeral, scalable runners. This required extending Forgejo's capabilities to support GitHub-compatible runner registration and webhook events.

Development Methodology & Contributions

Workflow

• Branching Strategy: Trunk-based development was utilized to ensure rapid integration.
• Collaboration: The team adopted Mob Programming. This practice proved essential for knowledge sharing and onboarding junior developers, creating a resilient and high-intensity learning environment.
• Versions: The platform evolved from Forgejo v7/8 to the current v11.0.3-edp1. An upgrade is pending to leverage the latest upstream GARM features.

Open Source Contributions

We actively contributed our extensions back to the upstream Forgejo project in a list of Codeberg.org pull requests

Artifact Caching (Pull-Through Proxy)

We implemented a feature allowing Forgejo to act as a pull-through proxy for remote container registries, optimizing bandwidth and build speeds.

• Source Code Branch: refactor-remote-registry-client

Key Performance Indicators (KPIs)

These KPIs measure the effectiveness of the Forgejo setup and quantify our strategic commitment to the Forgejo community.

KPI	Description	Target / Benchmark
Deployment Frequency	Frequency of successful pipeline executions.	High (Daily/On-demand)
Artifact Cache Hit Rate	Percentage of build requests served by the local Forgejo proxy.	> 90% (Reduced external traffic)
Upstream Contribution	Percentage of GARM-related features contributed back to Codeberg.	100% (No vendor lock-in)
PR Resolution Time	Average time for upstream community review and merge.	< 14 days (Healthy collaboration)