π
deploy-code-server A collection of one-click buttons and scripts for deploying code-server to various cloud hosting platforms. The fastest way to get a code-server environment!
Name | Type | Lowest-Price Plan | Deploy | |
---|---|---|---|---|
Coder | Anything | Open Source | read the docs | |
AWS EC2 | VM | Free Tier, 1 CPU, 1 GB RAM | see guide | |
DigitalOcean | VM | $5/mo, 1 CPU, 1 GB RAM | see guide | |
Vultr | VM | $3.50/mo, 1 CPU, 512 MB RAM | coming soon | |
Linode | VM | $5/mo, 1 CPU, 1 GB RAM | see guide | |
Railway | Container | Free, Shared CPU, 1 GB RAM |
see guide | |
Heroku | Container | Free, 1 CPU, 512 MB RAM | see guide | |
Azure App Service | Container | Free, 1 CPU, 1 GB RAM | see guide | |
Oracle Cloud | Terraform / OCI | Free Tier Support | see guide |
code-server on a VM vs. a Container
- VMs are deployed once, and then can be modified to install new software
- You need to save "snapshots" to use your latest images
- Storage is always persistent, and you can usually add extra volumes
- VMs can support many workloads, such as running Docker or Kubernetes clusters
π Docs for the VM install script
- Deployed containers do not persist, and are often rebuilt
- Containers can shut down when you are not using them, saving you money
- All software and dependencies need to be defined in the
Dockerfile
or install script so they aren't destroyed on a rebuild. This is great if you want to have a new, clean environment every time you code - Most app platforms do not support running docker or virtual volume mounts in the container.
- Storage may not be persistent. You may have to use rclone to store your filesystem on a cloud service. Documented below:
π Docs for code-server-deploy-container