Magma

Magma is a cool new distributed filesystem made for Linux. It uses libfuse on the client side and has its own special way of handling things on the server side with a distributed hash table.

How Does Magma Work?

So, Magma stores files, directories, devices, and even more by connecting a bunch of dedicated servers together. Each item stored in this network is called a flare. This could be anything like a file or a directory. To keep track of these flares, it uses the SHA1 hashing algorithm.

The Role of Nodes in Magma

The hash key for each flare is sent to the right server node, known as a vulcano. Each vulcano manages part of the hash key space that sits between two keys. But here's the twist: this space can change over time! There's a load balancer that works to share the load across all nodes fairly. It takes into account available disk space, bandwidth, and how many actual keys are on each node. So sometimes a key might shift from one node to another.

Magma's Compatibility

This program aims to stick as closely as possible to POSIX semantics, which helps keep things running smoothly. They're even planning to add features like file locking down the line! One thing to note is that clients can only connect to their local magma server. This helps avoid issues with NAT or VPNs.

Status of Development

Magma is still in its early stages, so don’t expect everything to work perfectly just yet. Some parts are pretty solid—like the flare system and client-server protocol—while other features are either unstable or still being planned out.

If you're curious about trying it out, check out this link for more info: Download Magma!