To view this content in our official product documentation, click here.

description: >- Full self-hosting documentation is coming soon. This page provides a high-level overview of what's covered.

Self-hosting

Deploy Patchworks Core into a customer-managed Kubernetes environment using Helm. Self-hosting is designed for organisations that need greater control over their infrastructure, hosting location, networking, or security posture - while still using Patchworks Core to build and run integrations.

Overview

A self-hosted Patchworks deployment runs the full Patchworks application and its supporting services inside Kubernetes. Deployment is delivered using Helm charts, with a split chart approach recommended for most environments - separating infrastructure and application resources into patchworks-infra and patchworks-app.

Architecture

A self-hosted environment is made up of two layers. The infrastructure layer covers the baseline services Patchworks depends on - including MySQL, Redis, RabbitMQ, Elasticsearch, and object storage. The application layer covers the Patchworks application itself - gateway, web services, schedulers, workers, and migrations.

Prerequisites

Before deploying, you'll need a working Kubernetes cluster, Helm, kubectl, container registry access, and a plan for DNS, TLS, and secrets management. Local proof-of-concept deployments can use kind with all services running in-cluster.

Installation

Installation follows a structured sequence: prepare a shared values file, install the infrastructure chart and confirm services are healthy, then install the application chart and validate migrations, ingress, and dashboard access.

Configuration

Everything is configured through Helm values - covering application URLs, database and cache connections, message broker settings, object storage, ingress, worker configuration, and secrets.

Worker modes

Patchworks supports three worker deployment models - including microservice - each suited to different scaling requirements and operational complexity. The right mode is typically agreed during implementation.

Operational checks

After deployment, a standard set of checks should be completed across pod status, migration and seeder jobs, worker logs, queue health, storage access, search availability, ingress routing, and application login.

Production considerations

Running Patchworks in production requires attention to high availability, backups, monitoring, centralised logging, secrets management, worker scaling, security controls, upgrade processes, and disaster recovery planning.