IPFS

IPFS, the InterPlanetary File System, is a peer-to-peer distributed file system and content-addressed, versioned protocol intended to make the web faster, safer, and more open. Instead of locating data by its hosting location, IPFS uses content identifiers derived from the data itself, allowing retrieval from any node that holds the content. It is an open-source project under ongoing development by Protocol Labs and a global community of contributors.

Files added to IPFS are broken into blocks and addressed with content identifiers called CIDs. These blocks form a Merkle DAG, enabling the reconstruction of a file from its parts. Nodes exchange blocks using a BitSwap-based protocol over a libp2p network, and they discover peers via a distributed hash table. A mutable pointer to a changing file can be maintained with IPNS, which maps a key to the current root CID.

IPFS uses the UnixFS format for files, integrates with DHT-based peer discovery, and provides gateways to bridge IPFS content with the traditional HTTP web. Data can be pinned to prevent garbage collection, promoting persistence across the network. The IPFS ecosystem can interoperate with Filecoin for long-term, incentive-based storage, but IPFS itself remains usable as a standalone, content-addressed system for distribution and hosting.

Common use cases include decentralized website hosting, large-file distribution, data archiving, and resilient content delivery in environments with intermittent connectivity. Challenges include variable retrieval latency, dependency on network participation for data availability, complexity of pinning and garbage collection, and potential privacy concerns due to broad data replication.