---
title: Architecture
url: /docs/administration/self-hosting/operations/architecture/
---
> **Note:** Self-hosting is only available with **Pulumi Business Critical**. If you would like to evaluate the self-hosted Pulumi Cloud, [request a Proof of Concept (PoC)](/product/self-hosted/#self-hosted-trial) or [contact us](/contact/).

This page describes the high-level architecture of a self-hosted Pulumi Cloud deployment. For detailed configuration of individual components, see [Components](/docs/administration/self-hosting/components/).

## Core components

| Component | Description |
| :-- | :-- |
| API service | Go-based REST API that handles CLI requests, state management, and all backend operations |
| Console | Web UI served as a static Angular application |
| Database | MySQL 8.0.x for metadata, stack state references, and user/organization data |
| Workflow runners | Docker hosts or Kubernetes cluster that runs the workflow runners for Pulumi Deployments and Insights scanning (Optional) |

## Supporting infrastructure

| Component | Description |
| :-- | :-- |
| Object storage | Blob storage for checkpoint (state) files and policy packs. Supported: S3 and compatible implementations, Azure Blob Storage, Google Cloud Storage |
| Search | OpenSearch 2.x or Elasticsearch 7.x for resource search and AI features |

## Data flow

```mermaid
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

- [Database best practices](/docs/administration/self-hosting/operations/database/) - MySQL HA, sizing, and migration guidance
- [Compute sizing](/docs/administration/self-hosting/operations/compute-sizing/) - Container resource allocation by platform
- [Object storage](/docs/administration/self-hosting/operations/object-storage/) - Bucket architecture, versioning, and lifecycle
- [Networking](/docs/administration/self-hosting/operations/networking/) - Multi-AZ deployment, load balancing, and TLS