FAQ

Question#

Anything that makes this project better than Kamal in any particular scenario?

Answer#

I have a love hate relationship with Docker. For me it adds to the learning curve. I have used Docker at work. But we only used in development environment. Also it was setup by Devops team. Only time I had an issue was, it ran out of disk space after few months.

This cannot be compared with Kamal because Kamal is Capistrano for Containers. HiveGrid uses Capistrano to deploy the App. It does not use Docker. This is not a deployment tool. It will automatically provision a EC2 instance for your Rails with default stack that consists of Postgresql 16, Redis, Puma and Caddy. It does not impose any requirement for a load balancer to be used or Cloudfare to be used. You don’t need to Dockerize your Rails app. Right now, if you go to https://clickplan.net/up , you can see a green bar. All the steps are documented on https://www.hivegrid.dev

If you need the flexibility to choose what you want to install and manage your EC2 instance with monitoring tools provided by AWS, then it will be good choice for your project.

Kamal seems to be motivated by the ability to switch between different cloud providers. HiveGrid allows you to create an image for Digitial Ocean if you want and provision a server using Digital Ocean. Packer has providers for Digitial Ocean which does the heavy lifting. Any contributions to add support are welcome.

AI Revised Answer#

HiveGrid Advantages#

Simplicity and Out-of-the-Box Solution:

HiveGrid is beneficial for users who prefer a straightforward, less technical setup for deploying Rails applications. It automates the provisioning of EC2 instances with a default stack including Postgresql, Redis, Puma, and Caddy. This is ideal for users who want a ready-to-go solution without the complexities of containerization or additional infrastructure management.

No Docker Requirement:

For users with a preference or requirement to avoid Docker, either due to complexity or the learning curve, HiveGrid provides a clear advantage. It allows the deployment of applications without needing to dockerize them, simplifying the development and deployment process for those unfamiliar with Docker.

Customizability and Integration:

While HiveGrid does automate certain aspects, it still offers flexibility in terms of backend services (like opting out of using a load balancer or Cloudflare). It’s also straightforward to adapt to platforms like Digital Ocean using Packer, catering to users who might prefer or require these platforms.

Kamal Advantages#

Provider Agnosticism:

Kamal is designed to facilitate easy switching between different cloud providers, which can be an advantage in scenarios where deployment flexibility across various clouds is required. This makes Kamal suitable for multi-cloud strategies, reducing vendor lock-in.

Containerization Support:

For teams that are already invested in Docker and containerized environments, Kamal offers a streamlined workflow that integrates tightly with these technologies.

Conclusion#

HiveGrid might be better suited for projects where simplicity in deployment is prioritized, and where the team does not wish to engage deeply with container technologies. On the other hand, Kamal would be preferable in environments where flexibility in cloud provider choice and containerization are critical operational requirements.

Revised Answer#

HiveGrid does not need any Dockerfile or Docker to be installed on the server to ship your Rails app to production.

The term deployment is confusing because it has different meaning based on whether you are a developer or a Devops engineer. As a developer the term deployment means: Application code is pushed to production server. Your customers can use the features of your Rails app.

From a DevOps engineer perspective, the term deployment means:

Installing software that is needed for your application to run on production such as web server, application server, database server etc. Configuration of these servers that configures these software and networking are also part of the deployment. The concepts of VPC, Subnet, Internet Gateway, CIDR and other networking terms are used in the Terraform script to provide a virtual cloud for your use on AWS.

So we can say that the term that is broad, deployment, can specifically mean either application deployment or infrastructure deployment based on the context or perspective.

So Capistrano is a application deployment tool. HiveGrid is a infrastructure deployment tool.

The up endpoint is the Health Check endpoint for Rails. It means connection to Postgresql database is working. In my case, Caddy is setup as reverse proxy to Puma process, which means it is also working.

I agree that is good feature that Kamal provides you the flexibility to switch providers. It leverages Docker to provide this flexibility.

HiveGrid can also provide that flexibility, but in this case you have to write Packer template for that specific cloud provider. It is not like no code portability like Kamal.

AI Revised#

Key Advantages of HiveGrid Over Kamal in Specific Scenarios:

Simplicity and Ease of Use:

HiveGrid is particularly advantageous for users who prefer a straightforward deployment process without the complexities of container management. It does not require Docker or Dockerfiles, making it an excellent choice for teams with limited Docker expertise or those who prefer to avoid Docker’s operational overhead.

Pre-configured Environment:

HiveGrid automates the setup of a standard Rails application environment, including web, app, and database servers, directly on EC2 instances. This pre-configured setup reduces the time and technical knowledge required to get a Rails application up and running, as opposed to Kamal, which may require more detailed configuration and understanding of Docker.

Direct Server Management:

Unlike Kamal, which emphasizes containerization and might abstract some of the lower-level details of the infrastructure, HiveGrid provides direct access to the underlying servers. This can be beneficial for scenarios where direct server access and management are required for compliance, performance tuning, or specific configuration needs.

No Need for Load Balancers or Additional Networking Tools Initially: HiveGrid does not impose mandatory requirements for load balancers or tools like Cloudflare from the get-go. This can be beneficial for smaller projects or applications where such complexity is unnecessary, reducing both costs and setup time.

While Kamal offers excellent flexibility in switching between cloud providers and is ideal for containerized environments, HiveGrid shines in environments where a more traditional server setup is preferred, and ease of setup is crucial. It offers a less complex alternative for Rails deployments, particularly for teams new to DevOps practices or those who do not wish to engage with the intricacies of containerization.