Network and Virtualization

Networking

The main challenge for a decentralized cloud is to build a network that operates reliably even when most nodes are unavailable.

The Leea Protocol network uses a mesh topology, where each node is aware of other nodes, their locations, and their network bandwidth.

Dynamic routing, based on DSDV and AODV protocols, allows tracking when nodes join or leave the network and automatically rebuilds routes to ensure the current node remains accessible without reducing its bandwidth.

To ensure security, all traffic within the network is encrypted using the RSA algorithm. We use Fastd to create secure and fast tunnels between nodes.

At the same time, outgoing traffic (intended for services outside the Leea Protocol) utilizes the host's bandwidth directly to achieve maximum speed.

Thus, we provide maximum stability, security, and network performance.

Virtualization

Virtual machines are the foundation of any cloud infrastructure. It is crucial to build a reliable system where each VM is highly isolated from the node on which it runs. Equally important is using technologies that fully unlock the potential of the hardware executing the compute tasks.

Leea Protocol employs several technologies to achieve maximum performance without compromising security.

QEMU

QEMU is a rapidly evolving open-source project that powers the majority of VMs worldwide. Its advantages include near-native performance, incredible configuration flexibility, and support for a large number of operating systems. For instance, QEMU can run any operational system on Windows, Linux & Mac machines.

Apple Hypervisor Framework

Support for macOS VMs in QEMU is under development. To make macOS available in Leea Protocol, we use Apple Hypervisor Framework built into macOS

GPU Passthrough

GPU passthrough is a technology that allows the OS kernel to directly present an internal PCI GPU to a virtual machine. In simple terms, this enables a specific device, like a GPU, to be assigned to a VM and used without any performance degradation. Inside the VM, the GPU will appear as if it is physically connected.

The Leea agent will set up the necessary drivers for you, but you will need to manually enable IOMMU/VT-d in your BIOS settings. You can refer to your motherboard manual or check out this helpful article for more guidance.

Note: You cannot passthrough a GPU that is already being used as the main GPU. For example, if you are running Windows with your display connected to a GPU, you will likely need to use integrated graphics or a separate GPU for your monitor.

GPU Partitioning

GPU partitioning allows multiple virtual machines to share a single physical GPU.

Each VM receives a dedicated portion of the GPU rather than access to the entire device.

This feature utilizes the Single Root IO Virtualization (SR-IOV) interface, ensuring a hardware-backed security boundary and predictable performance for each VM.

VM Migration

The state of a VM is extremely valuable. Therefore, if a node is removed from the network, we automatically migrate the VM to another node, ensuring the integrity of the data within the VM. This process can also be initiated manually.