1. Docs
  2. Administration
  3. Self-Hosting
  4. Operations
  5. Architecture

Self-Hosted Architecture Overview

    Self-hosting is only available with Pulumi Business Critical. If you would like to evaluate the self-hosted Pulumi Cloud, sign up for the 30-day trial or contact us.

    This page describes the high-level architecture of a self-hosted Pulumi Cloud deployment. For detailed configuration of individual components, see Components.

    Core components

    ComponentDescription
    API serviceGo-based REST API that handles CLI requests, state management, and all backend operations
    ConsoleWeb UI served as a static Angular application
    DatabaseMySQL 8.0.x for metadata, stack state references, and user/organization data
    Workflow runnersDocker hosts or Kubernetes cluster that runs the workflow runners for Pulumi Deployments and Insights scanning (Optional)

    Supporting infrastructure

    ComponentDescription
    Object storageBlob storage for checkpoint (state) files and policy packs. Supported: S3 and compatible implementations, Azure Blob Storage, Google Cloud Storage
    SearchOpenSearch 2.x or Elasticsearch 7.x for resource search and AI features

    Data flow

      graph TB
    subgraph Clients
        CLI[Pulumi CLI]
        Browser[Web Browser]
    end

    subgraph Load Balancer
        LB[Application Load Balancer]
    end

    subgraph Compute - Stateless
        API[API Service]
        Console[Console]
    end

    subgraph Data - Stateful
        DB[(MySQL 8.0)]
        S3[Object Storage]
    end

    subgraph Search
        OpenSearch[OpenSearch]
    end

    CLI -->|api.domain| LB
    Browser -->|app.domain| LB
    LB --> API
    LB --> Console
    API --> DB
    API --> S3
    API --> OpenSearch
    Console -->|API calls| API

    The API and Console services are stateless - all persistent data lives in the database and object storage. This makes the compute tier straightforward to scale horizontally and recover from failures.

    Next steps