I spend more time than I probably should watching VPS offers on LET.
Most deals are easy to ignore. Either the specifications are too weak, the useful locations are unavailable, or the attractive price only covers the first few months.
GreenCloud’s 2026 Fourth of July sale was different.
The offer that caught my attention gives you nearly 8 GB of RAM, four EPYC CPU cores and almost 8 TB of monthly bandwidth for $74 paid once every three years. That works out to approximately $2.06 per month.
I bought one to run a collection of applications and self-hosted services through Coolify. Mostly I use it to run my Hermes agent which coolify manages for me.
You can view the available locations through here GreenCloud VPS. First navigate to the client area; and then under store we can see the offer.
The GreenCloud 7777 offer
The plan is called the 4th of July Special, although many people will probably recognise it as the 7777 deal because of the memory and bandwidth allocation.
At the time of writing, the configuration includes:
| Resource | Included |
|---|---|
| Memory | 7,777 MB RAM |
| CPU | 4 EPYC Milan cores |
| Storage | 47 GB NVMe RAID-10 |
| Bandwidth | 7,777 GB |
| Network port | 10 Gbps |
| IPv4 | 1 address |
| IPv6 | /64 allocation |
| Virtualisation panel | VirtFusion |
| Backup | 1 free backup or snapshot |
| Location | Chicago, Illinois DC2 |
| Term | Three years |
| Price | $74 |
The specifications vary slightly between locations. Some use EPYC Rome rather than Milan, while Asian locations may include less bandwidth. Check the configuration shown during checkout rather than assuming every location is identical.
The plan is also marked non-refundable, so inspect the location and specifications before paying.
You can check the remaining locations from the GreenCloud Fourth of July sale page.
Is it actually a good application server?
For roughly $2 per month, the CPU and memory allocation are unusually generous.
The 7.7 GB of RAM is enough to run several modest applications alongside PostgreSQL, Redis and the Coolify control panel. Four CPU cores also leave enough room for normal application builds, background workers and database activity.
There is one obvious limitation: 47 GB of storage is not much.
That is enough for application containers, small databases, logs and deployment artefacts, but it is not enough for a large media library, a busy image-hosting service or years of unchecked Docker images.
I would use this server for:
- Personal applications and internal tools
- Small SaaS products
- Go, Python, Node.js and PHP services
- Staging and preview deployments
- API servers and background workers
- Small PostgreSQL, MySQL and Redis instances
- Monitoring tools such as Uptime Kuma
- Lightweight self-hosted services
I would not use it as the only storage location for important files or a rapidly growing production database.
Set up off-server backups and keep an eye on disk usage. The included snapshot is useful, but it should not be your only backup.
Why install Coolify?
You could manage the server using Docker Compose, systemd units and a manually configured reverse proxy.
That works, but it becomes repetitive once you have more than a few applications.
Each application needs some combination of:
- A Dockerfile or Compose file
- Environment variables
- Domain routing
- HTTPS certificates
- Database credentials
- Persistent volumes
- Deployment hooks
- Health checks
- Backups
Coolify puts those tasks behind a web interface. It is an open-source, self-hosted platform for deploying applications and databases to infrastructure you control.
The experience is closer to Railway, Render, Heroku or Vercel than to manually administering Docker containers.
You connect a Git repository, configure the application, assign a domain and deploy it. Coolify handles the container build, reverse-proxy configuration and TLS certificate.
The difference is that the server remains yours.
There is no per-service fee, per-seat charge or usage-based application bill. Your main infrastructure cost is the VPS, although you are also responsible for updates, security, monitoring and backups.
That trade-off makes sense for personal projects, side projects and small services where a fully managed platform would cost much more than the application itself.
Step 1: Buy and provision the server
Open the GreenCloud VPS sale and find one of the Fourth of July Special plans. Go to client area from the above URL (on top right) then you will be able to see this page. Go here on the Holiday Sale.
If by the time you arrive; there is no Holiday sale then it means that the sale was over. Only few hrs remains now; so do hurry if you plan to purchase it.
Choose a location close to your users. For a general-purpose deployment in the United States, the NYC EPYC Milan option is a reasonable choice when it is available.
During provisioning:
- Select Ubuntu 24.04 LTS or Ubuntu 22.04 LTS.
- Set a strong root password or add an SSH key.
- Save the server’s IPv4 address.
- Wait for the VPS to finish provisioning.
I prefer Ubuntu 24.04 LTS for a new installation. Coolify’s automated installer supports Ubuntu 20.04, 22.04 and 24.04 LTS.
Step 2: Connect over SSH
From your computer, connect to the server:
ssh root@YOUR_SERVER_IP
Replace YOUR_SERVER_IP with the IPv4 address shown in VirtFusion.
Update the operating system before installing anything:
apt update
apt upgrade -y
Reboot if the upgrade installs a new kernel:
reboot
Wait for the server to return, then reconnect over SSH.
Step 3: Check your firewall
A self-hosted Coolify installation needs several ports during initial setup:
| Port | Purpose |
|---|---|
| 22 | SSH access |
| 80 | HTTP and certificate validation |
| 443 | HTTPS traffic |
| 8000 | Initial Coolify dashboard |
| 6001 | Real-time communication |
| 6002 | Browser terminal access |
Restrict SSH to your own IP address where practical.
After assigning a custom domain to the Coolify dashboard, you can close direct public access to ports 8000, 6001 and 6002. Ports 80 and 443 need to remain available for hosted applications and certificate issuance.
Docker modifies iptables directly, so do not assume a basic UFW configuration blocks every published container port. Review the official Coolify firewall documentation before tightening the server.
Step 4: Install Coolify
Run the current official installation command:
curl -fsSL https://cdn.coollabs.io/coolify/install.sh | bash
When running from a non-root account, use:
curl -fsSL https://cdn.coollabs.io/coolify/install.sh | sudo bash
The installer handles the main setup work, including:
- Installing Docker Engine
- Configuring Docker logging
- Creating Coolify’s directories
- Generating the required SSH keys
- Installing the Coolify containers
- Starting the management interface
The official installation guide is available in the Coolify documentation.
Do not install Docker through Snap before running the installer. Coolify does not support Snap-based Docker installations.
Step 5: Create the administrator account immediately
Once installation finishes, the script should display an address similar to:
http://YOUR_SERVER_IP:8000
Open it in your browser and create the first administrator account immediately.
The first visitor to the registration page can create the initial administrator, so leaving the page exposed without completing registration is a security risk.
Step 6: Give the Coolify dashboard a domain
Using an IP address and port is fine during installation, but the dashboard should have a proper domain.
Create an A record through your DNS provider:
coolify.example.com -> YOUR_SERVER_IP
Then add the domain to your Coolify instance settings.
Once the proxy and certificate are active, you should be able to open:
https://coolify.example.com
At that point, remove unnecessary public access to the initial dashboard ports where your network configuration permits it.
For application deployments, you can create separate records:
api.example.com -> YOUR_SERVER_IP
app.example.com -> YOUR_SERVER_IP
status.example.com -> YOUR_SERVER_IP
A wildcard record is more convenient when you expect to deploy many applications:
*.apps.example.com -> YOUR_SERVER_IP
You can then use domains such as:
project-one.apps.example.com
project-two.apps.example.com
uptime.apps.example.com
Step 7: Deploy your first application
In the Coolify dashboard:
- Create a new project.
- Add an application resource.
- Connect GitHub, GitLab or another supported Git source.
- Select the repository and branch.
- Choose the appropriate build method.
- Add environment variables.
- Assign a domain.
- Press Deploy.
Coolify can deploy applications using a Dockerfile, Docker Compose, Nixpacks or framework-specific build detection.
A Go application with a Dockerfile works well on this server because the final container can be small and use very little memory.
The same server can also host Python, Django, FastAPI, Node.js, Next.js, PHP and static applications.
Adding PostgreSQL or Redis
Coolify can create databases from the same dashboard.
For example, you can add:
- PostgreSQL
- MySQL
- MariaDB
- MongoDB
- Redis
The database can remain private to applications running inside Coolify’s Docker network. Avoid publishing database ports to the public internet unless you genuinely need remote access.
For a small deployment, I would reserve the server roughly as follows:
- 1 to 2 GB for Coolify, Docker and the operating system
- 1 to 2 GB for PostgreSQL
- A few hundred megabytes for Redis
- The remaining memory for applications and builds
These are not hard limits. Actual usage depends on the application.
Container builds can temporarily use a large amount of memory and CPU, particularly for Node.js applications. Avoid deploying several heavy builds simultaneously.
Configure backups before adding important data
The VPS includes one backup or snapshot, but a snapshot alone is not a complete backup strategy.
A provider-side snapshot can disappear with the account, suffer from the same provider outage or capture a database in an inconsistent state.
For important applications:
- Configure scheduled database backups in Coolify.
- Send backups to S3-compatible object storage.
- Keep at least one copy outside GreenCloud.
- Test restoring the backup.
- Set a retention policy so old backups do not consume all available storage.
The 47 GB disk is the first resource you are likely to exhaust.
Docker images, build caches and application logs accumulate quietly. Enable automated Docker cleanup and monitor disk usage with:
df -h
To see Docker’s storage usage:
docker system df
Do not run docker system prune blindly on a production server. Check what Docker intends to remove first.
Is this really a Railway replacement?
For my use case, it covers much of the same workflow:
- Git-based deployments
- Automatic container builds
- Environment-variable management
- Domains and HTTPS
- Application logs
- Databases
- Persistent storage
- Deployment rollbacks
- One-click services
It is not a managed platform in the same sense as Railway.
GreenCloud maintains the underlying physical infrastructure, but you remain responsible for the operating system and everything running inside it. You need to handle security patches, monitoring, backup verification and capacity planning.
There is also no automatic migration if the server fails. Your recovery process depends on your backups and documentation.
That is the price of running a platform for approximately $2 per month rather than paying separately for every application and database.
For a bank, a large commercial product or a workload requiring formal service guarantees, I would build a more redundant setup.
For personal tools, experiments, staging environments and small products, the value is difficult to ignore.
The final result
For $74 every three years, this setup gives me:
- Nearly 8 GB of RAM
- Four EPYC CPU cores
- A public IPv4 address
- Almost 8 TB of bandwidth
- A 10 Gbps network port
- Automated application deployments
- Managed HTTPS certificates
- Private databases
- Git-based updates
- A dashboard for applications and services
The entire software layer is open source, and the applications remain on infrastructure I control.
That is a strong foundation for running side projects without collecting another pile of monthly platform subscriptions.
The limited storage means it is not suitable for everything, and the non-refundable three-year payment requires some trust in the provider. My own experience with a separate GreenCloud production server has been stable for more than three years, which made that commitment easier for me.
Stock changes quickly during these promotions. Check the specifications and available locations before ordering.
View the GreenCloud VPS Fourth of July offers.
Affiliate disclosure: This article contains affiliate links. I may receive a commission if you purchase a server through these links, at no additional cost to you. I have also been running multiple separate production server with GreenCloud for more than three years, so this recommendation is based on my own experience and the insane new deal they have launched.
Top comments (0)