> ## Documentation Index
> Fetch the complete documentation index at: https://qovery-docs-new-messaging-and-agent-tasks.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Migrate to Kubernetes

> Leave the PaaS, keep the workflow. Migrate from Heroku, Render, or Railway to Kubernetes on your own cloud, fully automated by AI

## Overview

Move off Heroku, Render, or Railway onto infrastructure you own. The migration runs alongside your current platform and does not require months of planning, rewriting deployment configs, learning Helm charts, or re-architecting your stack.

Qovery removes that complexity entirely. One command: an AI agent reads your repository and deploys it into the cloud account you already own (AWS, GCP, Azure, or Scaleway), after you approve the plan. It analyzes your codebase, generates Dockerfiles, provisions databases, sets up environment variables, and deploys everything as standard Terraform and Kubernetes manifests. No re-architecture, no Kubernetes expertise required.

The migration is non-destructive: your new environment runs alongside your current platform. Switch your domain over only once you're confident. Stop paying for your old PaaS and everything keeps running, on infrastructure and cloud billing you fully own.

<Info>
  For complex migrations (microservices architectures, stateful workloads, compliance requirements), Qovery provides **cloud architect assistance**. [Talk to a migration engineer](https://www.qovery.com/book-a-demo) to get a personalized migration plan.
</Info>

## Why Migrate with Qovery?

<CardGroup cols={3}>
  <Card title="AI-Powered Migration" icon="wand-magic-sparkles">
    The `/qovery-onboard` AI skill maps your current platform's concepts to Qovery and guides your cluster setup, then `qovery-deploy` analyzes your codebase, creates Dockerfiles, provisions databases, and deploys - all from a single prompt
  </Card>

  <Card title="No Kubernetes Expertise" icon="graduation-cap">
    Qovery abstracts the complexity of Kubernetes. You deploy applications - Qovery handles pods, ingress, TLS, RBAC, and networking
  </Card>

  <Card title="Your Cloud, Your Control" icon="cloud">
    Everything runs on your own AWS, GCP, Azure, or Scaleway account. No vendor lock-in, no shared infrastructure
  </Card>

  <Card title="Cloud Architect Assistance" icon="headset">
    Qovery's team helps with complex migrations. [Book a demo](https://www.qovery.com/book-a-demo) to get a personalized migration plan
  </Card>

  <Card title="Compliance Ready" icon="shield-check">
    SOC2 Type II, ISO 27001, DORA, HIPAA, GDPR compliant infrastructure from day one
  </Card>

  <Card title="Cost Savings" icon="dollar-sign">
    Right-sized from the start. No over-provisioning, no surprise bills. Customers report \$200K+ annual savings vs. DIY Kubernetes
  </Card>
</CardGroup>

## Migrate from Any Platform

<CardGroup cols={3}>
  <Card title="Heroku" icon="h">
    Dynos, add-ons, Heroku Postgres, Redis - all mapped to Kubernetes equivalents automatically
  </Card>

  <Card title="Render" icon="server">
    Web services, background workers, cron jobs, managed databases - migrated with full parity
  </Card>

  <Card title="Railway" icon="train">
    Services, plugins, and managed databases - mapped to Kubernetes equivalents automatically
  </Card>

  <Card title="OpenShift" icon="redhat">
    DeploymentConfigs, Routes, BuildConfigs - converted to standard Kubernetes resources
  </Card>

  <Card title="VMs / Bare Metal" icon="hard-drive">
    Legacy applications running on EC2, DigitalOcean droplets, or bare metal servers
  </Card>

  <Card title="Docker Compose" icon="docker">
    Multi-container setups defined in docker-compose.yml - each service becomes a Qovery application
  </Card>

  <Card title="Other PaaS" icon="cloud-arrow-up">
    Fly.io, DigitalOcean App Platform, Google App Engine, Azure App Service, and more
  </Card>
</CardGroup>

## How It Works

```mermaid theme={null}
flowchart LR
    SRC[Your Current Platform] -->|AI analyzes| AGENT[Qovery AI Agent]
    AGENT -->|generates| DOCKER[Dockerfiles]
    AGENT -->|provisions| DB[Databases]
    AGENT -->|configures| ENV[Env Vars & Secrets]
    AGENT -->|deploys to| K8S[Kubernetes on Your Cloud]
```

1. **Install** the Qovery AI skill (one command)
2. **Ask** the AI agent to migrate your project
3. **AI analyzes** your codebase - detects languages, frameworks, databases, dependencies
4. **AI generates** optimized Dockerfiles if missing
5. **AI provisions** databases (PostgreSQL, Redis, MongoDB, MySQL) and configures environment variables
6. **AI deploys** everything to Kubernetes on your cloud provider
7. **You verify** and go live

## Getting Started

### Option 1: AI-Powered Migration (Recommended)

The fastest path. The AI agent handles the entire migration end-to-end.

<Steps>
  <Step title="Install the Qovery AI Skill">
    One command installs the skill globally for all your projects:

    ```bash theme={null}
    curl -fsSL https://skill.qovery.com/install.sh | bash
    ```
  </Step>

  <Step title="Open Your AI Coding Tool">
    Launch [OpenCode](https://opencode.ai), Claude Code, Cursor, VS Code Copilot, or any compatible AI coding tool.
  </Step>

  <Step title="Ask the Agent to Migrate">
    ```
    Migrate me from Heroku to AWS
    ```

    Or use the skill command directly:

    ```
    /qovery-onboard
    ```

    The `qovery-onboard` skill acts as your personal cloud architect:

    * **Understand** your context - platform you're leaving, cloud provider you're moving to, constraints
    * **Map** concepts from your current platform (dynos, add-ons, services...) to their Qovery equivalents
    * **Guide** you through cluster and environment setup

    From there, `qovery-deploy` takes over to get your code running:

    * **Analyze** your codebase and detect languages, frameworks, and databases
    * **Create** optimized multi-stage Dockerfiles if your project doesn't have one
    * **Provision** databases (PostgreSQL, Redis, MongoDB, MySQL) - managed or containerized
    * **Set up** environment variables and secrets
    * **Deploy** everything to Kubernetes on your cloud provider
    * **Watch** the deployment and auto-fix any issues (up to 3 retries per service)

    <Tip>
      The AI agent supports Node.js, Python, Go, Java, Ruby, PHP, .NET, React, Vite, Next.js, Rails, Django, Spring, Laravel, and many more frameworks.
    </Tip>
  </Step>

  <Step title="Migrate Environment Variables">
    The AI agent handles environment variable migration as part of the deployment process.

    For **Heroku** specifically, you can also export and import manually using the Qovery CLI:

    ```bash theme={null}
    qovery auth
    qovery context set
    heroku config --app <your_heroku_app_name> --json | \
      qovery env parse --heroku-json > heroku.env && \
      qovery env import heroku.env && \
      rm heroku.env
    ```

    <Warning>
      Import sensitive data (API keys, credentials, tokens) as **Secrets**, not Environment Variables.
    </Warning>
  </Step>

  <Step title="Verify and Go Live">
    Once the deployment is complete:

    1. Check your application status in the [Qovery Console](https://console.qovery.com)
    2. Test your endpoints using the auto-generated Qovery URL
    3. [Configure your custom domain](/configuration/application#domains) when ready
    4. Update your DNS records to point to Qovery
  </Step>
</Steps>

### Option 2: Manual Migration

For teams that prefer full control over every step.

<Steps>
  <Step title="Create Your Dockerfile">
    If your project doesn't have a Dockerfile, create one. Qovery provides templates for all major frameworks.

    Example for a Node.js application:

    ```dockerfile theme={null}
    FROM node:20-alpine AS builder
    WORKDIR /app
    COPY package*.json ./
    RUN npm ci
    COPY . .
    RUN npm run build

    FROM node:20-alpine
    WORKDIR /app
    COPY --from=builder /app/dist ./dist
    COPY --from=builder /app/node_modules ./node_modules
    EXPOSE 3000
    CMD ["node", "dist/index.js"]
    ```

    <Tip>
      Use multi-stage builds to keep your images small and secure. The AI agent generates these automatically when you use Option 1.
    </Tip>
  </Step>

  <Step title="Create Resources on Qovery">
    1. Sign in to the [Qovery Console](https://console.qovery.com)
    2. Create your Organization and Project
    3. Create your production Environment
    4. Add your application (select your Git repository, branch, and port)
    5. Add your database (PostgreSQL, MySQL, MongoDB, or Redis - managed or containerized)
  </Step>

  <Step title="Configure Environment Variables">
    1. Export your environment variables from your current platform
    2. Import them into Qovery via the Console or [CLI](/cli/commands/env)
    3. Create environment variable aliases to connect your services (e.g., link your app to your database using `DATABASE_URL`)
  </Step>

  <Step title="Deploy">
    Click "Deploy" in the Qovery Console or use the CLI:

    ```bash theme={null}
    qovery deploy
    ```
  </Step>
</Steps>

### Need Help? Talk to a Cloud Architect

<CardGroup cols={1}>
  <Card title="Book a Migration Demo" icon="calendar" href="https://www.qovery.com/book-a-demo">
    For complex migrations - microservices, stateful workloads, compliance requirements, or large-scale infrastructure - Qovery provides personalized cloud architect assistance. Get a migration plan tailored to your stack.
  </Card>
</CardGroup>

## What Gets Migrated?

| Source                                     | Qovery Equivalent                                                          | Handled by AI?             |
| ------------------------------------------ | -------------------------------------------------------------------------- | -------------------------- |
| Web applications (any language/framework)  | [Applications](/configuration/application) - containerized on Kubernetes   | Yes                        |
| Background workers (Sidekiq, Celery, etc.) | [Applications](/configuration/application) - separate services             | Yes                        |
| Cron jobs                                  | [Cron Jobs](/configuration/cronjob)                                        | Yes                        |
| PostgreSQL, MySQL                          | [Managed Database](/configuration/database) (AWS RDS, GCP Cloud SQL, etc.) | Yes                        |
| Redis, MongoDB                             | [Container Database](/configuration/database) or Managed                   | Yes                        |
| Environment variables                      | [Environment Variables & Secrets](/configuration/environment-variable)     | Yes                        |
| Custom domains                             | [Custom Domains](/configuration/application#domains)                       | Manual DNS update required |
| Add-ons (Memcached, Elasticsearch, etc.)   | [Helm Charts](/configuration/helm) or container services                   | Partially                  |
| Scheduled tasks                            | [Lifecycle Jobs](/configuration/lifecycle-job)                             | Yes                        |

## FAQ

<AccordionGroup>
  <Accordion title="How long does migration take?">
    For most applications, less than a day. The AI agent handles Dockerfile creation, database provisioning, environment variable setup, and deployment in minutes. Complex multi-service architectures may take longer - [book a demo](https://www.qovery.com/book-a-demo) for a personalized timeline.
  </Accordion>

  <Accordion title="Do I need Kubernetes expertise?">
    No. Qovery abstracts all Kubernetes complexity. You interact with applications, databases, and environment variables - not pods, deployments, or ingress controllers. The AI agent handles the Kubernetes-specific configuration automatically.
  </Accordion>

  <Accordion title="Which cloud providers are supported?">
    AWS (EKS), Google Cloud (GKE), Azure (AKS), and Scaleway (Kapsule). You can also bring your own Kubernetes cluster (BYOK).
  </Accordion>

  <Accordion title="What about data migration?">
    Database schema and application code are migrated automatically. For data transfer (copying production data), you'll need to handle the database dump/restore process. Qovery databases support standard import tools (pg\_restore, mysql import, mongorestore).
  </Accordion>

  <Accordion title="How to create a custom domain?">
    Check the [custom domain configuration documentation](/configuration/application#domains) for setup instructions.
  </Accordion>

  <Accordion title="How to monitor my apps?">
    Qovery provides built-in [observability](/getting-started/guides/qovery-101/observe) including logs, metrics, and alerts. You can also integrate with Datadog, New Relic, or other monitoring tools.
  </Accordion>

  <Accordion title="Do you have preview environments?">
    Yes. Qovery provides [Preview Environments](/getting-started/guides/use-cases/preview-environments) that automatically create a clone of your environment on every pull request.
  </Accordion>

  <Accordion title="How to rollback?">
    App rollback is available through the [deployment actions](/configuration/deployment/overview#rollback). You can roll back to any previous deployment in one click.
  </Accordion>

  <Accordion title="How does auto-scaling work?">
    Auto-scaling is configured in the [application settings](/configuration/application#auto-scaling). Qovery supports horizontal pod autoscaling based on CPU/memory usage.
  </Accordion>

  <Accordion title="Can I use Terraform?">
    Yes. Qovery has a [Terraform provider](/terraform/overview) for infrastructure-as-code deployment. The AI agent can also deploy via Terraform if you choose that option during setup.
  </Accordion>

  <Accordion title="Is it possible to get a shell into my app?">
    Yes, via the cloud shell button in the Qovery Console or through the CLI with `qovery shell`.
  </Accordion>

  <Accordion title="How to connect to external services (MongoDB Atlas, AWS services)?">
    Use [VPC peering](/configuration/clusters#vpc-peering) for secure, private connectivity to external services in your cloud account.
  </Accordion>
</AccordionGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="Deploy with AI Agent" icon="wand-magic-sparkles" href="/getting-started/quickstart/ai-agent">
    Full guide on installing and using Qovery AI skills
  </Card>

  <Card title="Production Setup" icon="server" href="/getting-started/guides/use-cases/production-environment-management">
    Set up a production-ready environment with best practices
  </Card>

  <Card title="Preview Environments" icon="code-branch" href="/getting-started/guides/use-cases/preview-environments">
    Automatically create environments for every pull request
  </Card>

  <Card title="Book a Demo" icon="calendar" href="https://www.qovery.com/book-a-demo">
    Get cloud architect assistance for your migration
  </Card>
</CardGroup>