Recommended Specifications
AOS is a lightweight Python program, so it does not take very much to run. As of AOS 5.4.1, most of our instances operate within the realm of 50-60mb memory and 200-350mb disk space.
Because of that, we recommend the following to be reserved for just AOS and a lightweight operating system (arch) in applications such as Proxmox LXC containers:
| Resource | Recommended | Minimum | Notes |
|---|---|---|---|
| CPU Threads | 2 | 1 | This will greatly depend on your expected traffic. If your AOS is clustered, you won't need as much. |
| Memory | 1GB | 512MB | Depends on your host OS, AOS Plugins, and server lifespan |
| Disk space | 6GB | 4GB | Depends on your host & log retention. AOS will not use over 512mb by itself. |
| Network | 100MB/s | N/A | All official AOS nodes have gigabit uplinks. Try not to throttle machines if possible. |
Please note that you may need to increase resource availability during the install process to ensure uv can write packages in a timely manner.
These are just base recommendations and are not all-encompassing, your needs will strongly vary (hosting mongo? heavy OS (ubuntu)? LXC, qemu, or bare metal? shared machine?). On us-1, AOS along with Arch Linux is operating with 200MB of memory and in 3GB of disk space roughly. These recommendations are just general safe assumptions based on our experience with 5 nodes.
At the moment, Docker is not supported out of the box, however, that does not mean you are unable to run it in a generic Python container. We are looking into docker support for the long term, but for now we are continuing to support systemd-based installs.