Migrate from Heroku

Northflank supports all your applications you've been deploying to Heroku, and migrating is straightforward and easy. You can use our automated import tool, or move your applications over manually.

You can also manage your account and projects through the Northflank API, CLI, and JavaScript client, helping you develop your applications locally using forwarding and command execute, and programmatically manage your entire devops workflow.

Our pricing is straightforward and transparent, and you only pay for the resources you use.

1. Create an account
2. If you're working with others, create a team and invite your colleagues
3. Link an account from a supported Git service to Northflank to build from your repositories

To run an automated import, navigate to your account dashboard and select the Heroku import option from the settings page.

Link your Heroku account to view your applications and pipelines in the Northflank importer, and select the ones you want to import. Northflank will automatically select a project to import to, change this if you'd like the resource to be imported somewhere else. You can import to a free project, providing the free project's limits are not exceeded.

Most options will be automatically configured from your Heroku settings, however you should check each field before importing, and complete or update where required.

• Resource names: as there is not always a 1-to-1 relation between Northflank and Heroku resources, some names will be automatically generated for your new resources
• Buildpack stack: Northflank will detect and use your buildpack configuration for each of your applications as a custom buildpack stack. You should check this is done correctly under advanced build options, especially if you are using legacy or custom buildpack stacks.
• Branch: the main branch will be automatically selected if you're using Heroku Git, or if your repository only has one branch. If your application is built from a repository with multiple branches you must select one.
• Environment variables: your secrets, including those provided by linked Heroku addons, will be copied to build and deployment services directly, or into secret groups. Northflank handles build arguments and runtime variables separately. Check the guides below to learn more about specific migration scenarios.

You cannot import Heroku applications with no dyno formations, apps from pipelines that span multiple regions (e.g. one app deployed in US, one in Europe), and you cannot import apps running images from the Heroku container registry.

Import guides for specific Heroku scenarios

Automated import

Northflank will import your single-dyno application as a combined service. The combined service will build your application, deploy the built image, and contain your environment variables.

Your secrets will be copied to the build arguments and runtime variables in the combined service. These secrets are separate, so if you edit the value of a build argument it will not affect the value of the runtime variable, and vice versa.

Manual import

To import a Heroku application with a single dyno formation (an application that contains dynos of only one process type, e.g. web or worker dynos), you can create a combined service.

All aspects of a combined service can be configured after creation - except the name.

1. Choose service from the create new menu in the top right corner of the dashboard
2. Basic information: select combined service and enter a name
3. Repository: select the Git repository from the drop-down list and choose the branch you want to build from
4. Build options: choose buildpack to automatically build your application. Northflank uses the Heroku 20 stack by default. You can select another buildpack stack, or add custom buildpacks, in advanced build settings.
5. Environment variables (optional): you can set build arguments and runtime variables, or add secret files in advanced. You can add your connection details for hosted databases or API keys for other services here to make them available within your application.
6. Networking (optional): Northflank will add a public port for buildpacks, you can redefine this port and add more if required. You can also create a service with no public ports that means it can only be accessed by other resources within your project.
7. Resources: choose a plan with more CPU and memory available to your builds and instances, and increase the number of instances to deploy, if required
8. Advanced (optional): you can configure health checks and manage other advanced features

Learn more about combined services

A combined service is a self-contained pipeline that builds and deploys commits to a branch in a Git repository.

Combined services are highly configurable: you can manage how they build your application, CI/CD, networking and domains, health checks, and more.

You can also scale your combined services to meet your needs, increasing the CPU power and memory available to each running instance and increasing the overall number of running instances.

You can learn more about managing combined services in our getting started guide, or check the documentation for specific tasks you want to achieve.

Northflank allows flexible and secure private and public networking for services, jobs, databases and other addons. You can create secure, private connections within your project, or expose ports publicly to access your applications online.

You can expose multiple ports on every deployment, public and private, and HTTP, HTTP/2, Websockets, gRPC, TCP and UDP are all supported networking protocols.

Northflank uses scalable and highly performant load balancers to securely distribute external traffic to your applications.

Networking settings are accessed on the ports & DNS page on deployment and combined services, and on the settings page in the network section for databases and other addons.

Domains

Public ports can use either a free code.run domain, generated by Northflank, or your own custom domains. All domains are secured with a free, auto-generated TLS certificate.

You can add your own domains in your account dashboard, then link them to specific ports in your services. You can quickly and easily change what port or service they are linked with, without needing to re-verify or restart your services.

1. Add a domain.

2. Add a subdomain.

3. Link a domain to a port.

Learn more about networking on Northflank, or follow a step-by-step guide to add and link a domain to a port

To get started on Northflank you can use your existing databases and other service integrations by including their API keys, connection strings, or other secrets as environment variables. These will be included with the automated importer, or you can add them manually yourself.

Northflank offers several managed databases, including PostgreSQL, Redis, MongoDB, MySQL, and MinIO, that can be deployed with a click.

Select addon from the create new and choose a database to deploy.

After deploying a database you can import your data from the backups page by uploading a text dump or specifying the connection string of the database you wish to import from. You have several options to automatically or manually manage runtime migrations on Northflank.

You can also create persistent volumes to attach to your deployments, for any use cases not covered by our managed addons.

Find guides to deploy and use databases and volumes on Northflank

If you're using Heroku scheduler you can create cron jobs on Northflank that will run your applications at specific times, before shutting down.

You can deploy from a repository, which allows you to automatically build any new commits to the tracked branch and so always run the latest version of your job, or deploy an image from a Northflank build service or an external container registry.

You can run the job on a cron schedule, or if a new image is available.

You can create and configure a new job from the create new menu.

Learn more about jobs on Northflank

You can create and manage projects, builds, and deployments programmatically using the API, interact with Northflank using the CLI, and integrate the JavaScript client into your applications to create solutions for any use-case you can imagine.

You can use the CLI to forward resources for local development, and to execute commands in your running containers.

To get started, create an API token, or create a token template for your team to use.

Learn more about using Northflank programmatically

Automated import

Northflank will import your application with multiple dyno formations as a build service and multiple deployment services with an associated secret group. If your Heroku application contains dynos of different process types (e.g. web and worker dynos), a deployment service for each dyno type will be created. Your Heroku application's secrets will be copied to your build service as build arguments and to a secret group as runtime variables, to be inherited by your deployment services.

Manual import

To import a Heroku application with multiple dyno formations (an application that contains dynos with more than one process type, e.g. web and workers dynos), you are recommended to create a build service, multiple deployment services, and a secret group (if required).

Build service

You can create a build service linked to the Git repository for your application, and build commits to specific branches and pull requests by adding build rules.

All aspects of a build service can be configured after creation - except the name.

1. Choose service from the create new menu in the top right corner of the dashboard
2. Basic information: select build service and enter a name
3. Repository: select the Git repository from the drop-down list and choose the branch you want to build from. You can also set build rules to build specific branches or pull requests.
4. Build options: choose buildpack to automatically build your application. Northflank uses the Heroku 20 stack by default. You can select another buildpack stack, or add custom buildpacks, in advanced build settings. You can also set the build context, if it's not the root directory.
5. Environment variables (optional): you can set build arguments or add secret files in advanced
6. Resources: choose a plan with more CPU and memory available to your builds, if required

Learn more about building on Northflank.

Deployment services

The deployment services will run containers from the images provided by your build service. You can create separate deployment services for each type of dyno in your Heroku application. For example, you would create one deployment service for your web dynos, and one deployment service for your worker dynos. These can then be configured and scaled separately.

For web dynos you can add the required HTTP ports and publicly expose them to the internet. For worker dynos you can add any required HTTP, TCP, or UDP ports and they will only be accessible to resources within your project.

You can either add variables required at runtime directly in the deployment services on the environment page, or create a secret group that can be inherited by multiple services.

1. Choose service from the create new menu in the top right corner of the dashboard
2. Basic information: select deployment service and enter a name
3. Deployment: select Northflank and link your build service, selecting a branch to deploy builds from
4. Environment variables (optional): you can set runtime variables or add secret files in advanced
5. Networking (optional): add public ports to make the service accessible on the internet, or private ports for it to only be accessed by other resources within your project
6. Resources: choose a plan with more CPU and memory available to your instances and increase the number of instances to deploy, if required
7. Advanced (optional): you can configure health checks and manage other advanced features

Learn more about deploying on Northflank.

Secret group

Rather than adding secrets to each of your services individually, you can create secret groups that can be inherited by multiple services. On Northflank environment variables for build arguments and runtime variables are normally separate, however your can create a secret group containing both. Variables in secret groups are inherited by all your services by default, but you can restrict them to specific services if required.

1. Choose secret group from the create new menu in the top right corner of the dashboard
2. Basic information: enter a name and select the type of secrets the group will contain: runtime variables, build arguments, or both
3. Secrets: enter your secret keys and values using the key-value table edit view, as JSON, or upload or enter in ENV format. You can also add secret files in advanced.
4. Linked addons: you can link addons to your secret group to make them easily accessible in your services and jobs
5. Advanced: by default secret groups will be inherited by all your services and jobs in the project. You can choose to apply the secret group to specific services and jobs only.

Learn more about secrets on Northflank.

Automated import

Northflank will import each pipeline application as a build service and deployment services with associated secret groups. If your Heroku application contains dynos of different process types (e.g. web and worker dynos), a deployment service for each dyno type will be created.

The deployment services will be placed in staging and production stages in the pipeline, which you can then link together to configure your workflow.

The commit currently deployed to your Heroku apps, regardless of their stage in a pipeline, are built and deployed to your deployment services for that app during the import. See the guides on pipelines and CI/CD below for more information on managing your releases.

Each application's secrets will be copied to a build service as build arguments and to a secret group as runtime variables, to be inherited by your deployment services.

Manual import

To import your pipeline to Northflank manually you can create a build service for each repository you wish to build from, and set rules to build specific branches or pull requests.

You can then create deployment services, one for each type of dyno per application, to populate a pipeline.

You can add secrets as build arguments directly to build services, and runtime variables directly to deployment services, or create secret groups to contain environment variables that can be inherited by multiple services.

You can follow the steps in import a Heroku application with multiple dynos to Northflank to create build services, deployment services, and secret groups.

Pipelines on Northflank

Create a pipeline and link a build service

1. Choose pipeline from the create new menu in the top right corner of the dashboard
2. Enter a name and create pipeline
3. Drag your deployment services into the staging and production stages, or click the add button and select a stage
4. Select a build service with the repository you want to deploy builds from, and link either a branch or pull request that the service is configured to build from to a deployment service in staging
5. Your build service will automatically build from the tracked branches and pull requests, and your linked deployment service will automatically deploy the latest available build

Link deployment services in stages

1. Click the link button on the first deployment service in the chain you want to create. It will now be highlighted in green.
2. Click on the deployment service(s) in the next stage you want to link it to
3. Click again on the link button on the first deployment service you were linking from, or outside the deployment cards, to exit the link editor for that service

Promote builds

1. Click the promote button on the deployment service that is running the build you want to promote
2. Any linked deployment services in the next stage will be redeployed with the selected build. Continuous deployment will be disabled, so they will continue to run the selected build until a new one is promoted