# Sizes and scaling

## The six sizes

| Size  | vCPU | Memory | NVMe SSD |
| ----- | ---- | ------ | -------- |
| Pro 1 | 1    | 2 GB   | 20 GB    |
| Pro 2 | 2    | 4 GB   | 50 GB    |
| Pro 3 | 2    | 6 GB   | 80 GB    |
| Pro 4 | 4    | 8 GB   | 120 GB   |
| Pro 5 | 6    | 16 GB  | 240 GB   |
| Pro 6 | 8    | 24 GB  | 400 GB   |

Every size includes root access, the nightly backup, monitoring and Full-Stack Assistance. Prices, monthly or yearly with one month free, are on the [VPS Pro page](https://www.simafri.com/en/vps/).

## Choosing your size

A few pointers, to refine with our team according to your project:

| Your project                                             | Starting point |
| -------------------------------------------------------- | -------------- |
| a website, an API, a bot, a test environment             | Pro 1 or Pro 2 |
| a web application with its database, in production       | Pro 3 or Pro 4 |
| several applications, a large database, heavy processing | Pro 5 or Pro 6 |

Memory is often the deciding factor: count the memory of your database, of each process of your application, and the headroom a build requires (an `npm run build` or a `docker build` takes a lot, for as long as it lasts). Tell us what you want to run: we recommend the right size.

## Moving up to a larger size

Your server moves up to a larger size **at any time, without reinstalling**: your data, your addresses and your settings stay in place.

1. Write to support with your server's name and the size you want.
2. Our team sends you the proforma for the change, at the rate of the new size from the date of the change.
3. The memory and the processor adjust without a reboot, and the disk grows, without interrupting your services.

The new size shows straight away in your server's card, on the **Resources** line.

## Moving down to a smaller size

Write to support with the size you want: the change takes effect the following month, at your next due date.

## Tracking your resources

| Resource  | The command        | What it tells you                                        |
| --------- | ------------------ | -------------------------------------------------------- |
| memory    | `free -h`          | the memory used and the **available** memory             |
| processor | `htop` or `uptime` | the load, compared with the number of vCPUs in your size |
| disk      | `df -h /`          | the space used and the free space                        |

Our monitoring also tracks your server's load and disk space, and alerts you when an indicator calls for it: it is often the right moment to tidy up (`apt clean`, `journalctl --vacuum-size=200M`, old exports) or to move up to a larger size.

## Frequently asked questions

### My application runs out of memory during the build.

Build on your computer or in your continuous integration and deploy the result, or move up to a larger size. Your server's swap space absorbs brief peaks.

### What happens when my disk is full?

Writes fail, and a database stops. Our monitoring alerts you beforehand: free up space with `ncdu /` to find what is taking up the disk, or move up to a larger size, whose disk is bigger.