What makes IPFS different from other p2p file systems?

There is many distributed file system exist today (Hive2Hive, Ceph … ) what makes IPFS different ?

1 Like

Which p2p filesystems, specifically, are you interested in knowing the differences between? There are so many differences between even the two you listed it’s hard to know what you’re looking for in an answer.

For the two you listed:

Hive2Hive - I’d never heard of this before, but this looks like it’s primarily a file synchronization tool, not a p2p filesystem. This seems more like it’s trying to replace Dropbox (etc) and compete with other p2p synchronization tools like Resilio Sync or Syncthing.

Ceph - this isn’t a p2p filesystem either. It’s a distributed filesystem.

P.S. It’s generally not good practice to post the same question multiple times.

Ok, it seems like I dont see the difference between p2p file system and distributed file system

The main distinction in my mind is that every node in a p2p file system is essentially the same, whereas nodes in a non-p2p distributed file system are not equal and are often specialized. For example, LizardFS is a distributed file system but has master/shadow servers, chunkservers, and metaloggers (not sure if you’d count clients as a separate node type, but I would).

Distributed file system

Distributed filesystems generally allow the function of each piece of their infrastructure to be split across multiple machines for performance, scalability, or fault tolerance. What that looks like in practice depends on the specific file system implementation (e.g., Ceph, GlusterFS, LizardFS, MooseFS, etc).

Reference and more information here.

P2P filesystem

distributed systems in which all nodes have identical capabilities and responsibilities, and all communication is symmetric.

P2P systems would be a subset of distributed systems.

Definition copy/pasted from here.

2 Likes

So can we say that a p2p system is a distributed system with a “devolution of governance” ?

I’d agree with that.