Decentralized File Storage: Five Projects To Replace The Cloud
Share this article
Data generation is on overdrive, and consumer requirements are far outpacing the capacity of individual hard drives. Current estimates suggest that 2.5 quintillion bytes of data created every day, flooding out of numerous connected devices.
Most cloud storage solutions rely on systems like Dropbox, Microsoft OneDrive, Google Drive, and Apple iCloud Drive. These platforms are highly centralized, with single points of failure that are highly susceptible to breaches. Thus, they undermine privacy, keep the prices of data storage high, and are prone to errors resulting in costly Internet outages down the line for users.
Decentralized storage is now gaining traction as a secure solution. Starting with Interplanetary File Storage, several crypto projects are seeking new ways to secure user data.
These are the main players in decentralized file storage:
InterPlanetary File System
The IPFS Protocol is a P2P (peer-to-peer) distribution protocol that attempts to “connect all computing devices with the same system of files,” just like the documents in your hard drive. Although similar to the Web, IPFS is more similar to a BitTorrent swarm, which exchanges objects with a GitHub repository. Unlike the web, data on IPFS has no single point of failure, and nodes don’t need to trust one another.
IPFS in itself is becoming a primary subsystem of the internet, which could complement or even replace HTTP. The system combines ideas from previous P2P systems such as BitTorrent, Git, self-certified filesystems (SFS), and distributed hash tables into a single cohesive ecosystem, that distributes large data while allowing for writing and deploying applications.
Since the data is content-addressed, nodes don’t need to trust each other. Instead, they connect, transfer, and store objects (i.e. files and data structures) in local storage.
Here’s what happens when you add a file to the IPFS Protocol:
- Each data file is given a cryptographic hash, which functions as a unique fingerprint.
- IPFS removes duplications across the network and tracks the version history of each file.
- Each network node only stores what it is interested in, while indexing the information to help determine who is storing what.
- Anytime someone wants to look up a certain file, they ask the network to find nodes storing content associated with the file’s hash.
However, IPFS is not free of limitations, and the lack of concrete incentives cause a barrier for long-term file storage. Several projects are seeking to overcome these problems by allowing users to pay for storage with cryptocurrency.
Storj is the market leader of decentralized storage with a community of about 20,000 users and 19,000 storage providers. The project is currently in public alpha, and costs $0.015 per gigabyte of storage per month, according to the website.
In the Storj protocol, nodes are selected to host data based on ping time, throughput, and other criteria. Users will not have to choose a specific timeframe to store their files, as the data will remain there indefinitely until it is removed. The protocol utilizes sharding, encryption, and swarming models to split and recompile files.
Currently, Storj is “payment agnostic” but uses the ERC-20 token STORJ to compensate storage providers. In the future, the platform says it will offer payments through other cryptocurrencies, fiat currencies, credit, or “even physical transfer of live goats.”
When it comes to having a viable product, Sia is one of Storj’s top competitors. As of the time of writing, it costs $0.28 to store a terabyte of data for a month, with small fees for uploading and downloading data, as well as contract formation.
Sia maintains storage through smart contracts. Each file is split into thirty encrypted segments, but can be rebuilt using only ten of these segments. Records in the Sia network are encrypted, in the same manner as those on Storj.
Each host then promises to store data for a specific length of time, which can be cryptographically verified without exposing the data. The host is then compensated for every proof they submit, and penalized for missing proofs.
Sia has its own blockchain, which supports the smart contracts that are used to send and manage files, as well as making storage proofs to be publicly available and verifiable. Therefore, it needs miners to support its blockchain and rewards them in SiaCoin.
FileCoin is a project from Protocol Labs, the same company that has successfully deployed IPFS. It’s widely known for launching one of the most successful ICOs, which raised over $200 million in token sales. It also has backing from players such as Kamal Ravikant and Andreessen Horowitz.
With Filecoin, users are able to program how long they’d like to offer or provide services.
The FileCoin data service does not rely on a single coordinator, but rather a client, storage miner, and retrieval miner. Clients pay to store and retrieve data, storage miners earn tokens by offering storage, and retrieval miners earn tokens by delivering the data.
The mining power, or the amount of data stored per miner, is proportional to the amount of active storage. This is expected to incentivize “miners to amass as much storage as they can, and rent it out to clients.”
The decentralized storage provider states its cloud storage network provides security as stored content is encrypted end-to-end, and storage providers do not have access to encryption keys.
FileCoin is expected to launch in the middle of this year.
Swarm is a decentralized P2P storage and sharing service that stores all of its data in nodes, and operates as a native base layer of the Ethereum Web3 stack. The protocol seeks to cache “Ethereum’s public record, in particular to store and distribute dApp code and data as well as blockchain data.”
In order to attack the Swarm network, one would need access to all the nodes at the same time – which would require hundreds of supercomputers.
Swarm operates with an upload and disappear model, which allows users to upload content, which can then view it offline via synchronization. Synchronization is a procedure that allows nodes to continuously pass data and information between one another.
The protocol is still in its Proof of Concept stage and currently operates on the Ethereum TestNet.
A future in distributed storage networks?
Decentralized solutions may soon make an impact on the future of storage systems. Although these systems require several improvements before they can become stable, statistics reveal a growing global demand. Additionally, they are incredibly low-cost, transparent, and more secure than present systems.
However, decentralized cloud storage will not happen overnight, and many roadblocks can be expected before mass adoption.
The author is invested in Ethereum, which is mentioned in this article.