Beware of Centralized Storage in Decentralized Clothing
Disclaimer: The Industry Talk section features insights by crypto industry players and is not a part of the editorial content of Cryptonews.com.
All blockchain storage solutions are created equal. But some are more equal than others.
Decentralized storage services work in broadly the same manner, using InterPlanetary File System (IPFS) or a variant thereof. This distributed file system enables peer-to-peer sharing and storage of media and is designed to provide a decentralized alternative to HTTP for serving and retrieving content on the web.
The problem with such solutions is that they are only as decentralized as the servers used to hold the data. And in this respect, it turns out that blockchain developers face something of a quandary. On the one hand, they claim to love the notion of decentralization and the freedom to create innovative products that make use of IPFS. But on the other hand, they’re spoiled by the speed, availability, and low latency of web2 infrastructure.
This has led to some blockchain projects falling back on the crutch of centralized servers on which to host their “decentralized” data. It may sound oxymoronic but the practice is endemic throughout the industry. At one point, 70% of all Ethereum nodes were running on centralized infra with cloud providers such as Google, Amazon, and Microsoft doing the bulk of the heavy lifting. Understanding why this is a problem calls for a reprise of how decentralized storage works.
The Case for IPFS
In IPFS, files are addressed by their content and not their location, which means that each file is given a unique cryptographic hash that serves as its address. This allows files to be stored and retrieved from multiple nodes in the network, making them highly resilient to censorship and failure.
IPFS also allows for content addressing, which means that identical files will have the same address, regardless of who is hosting them. This enables efficient content distribution and reduces duplication of content across the network.
However, content duplication is exactly what blockchain nodes are designed to do: each one holds an identical copy of the network’s latest state. This provides redundancy, enables network consensus to be achieved, and ensures that collusion is virtually impossible.
When a significant number of nodes are stored on centralized servers like AWS, this doesn’t mean that Amazon could commandeer the network on a whim; there are still enough truly decentralized Ethereum nodes to prevent this from happening. However, this doesn’t disguise the fact that reliance on web2 infrastructure defeats the purpose of technology such as IPFS and impairs blockchain’s entire value proposition.
Use It or Lose It
Blockchain developers are fortunate to have access to so many decentralized storage solutions. Through projects like Filecoin and Arweave, access to decentralized infrastructure has never been greater. But if these solutions aren’t utilized properly, to store content on self-hosted servers rather than in the cloud, much of the benefits this technology can deliver will go to waste.
More importantly, projects that follow this approach are misleading their communities, who are being sold a lie – or at the very least, a deception. To misquote Benjamin Franklin, "Those who would give up essential freedom, to purchase a little temporary convenience, deserve neither freedom nor convenience."
Decentralized storage provides that freedom from censorship, surveillance, and security scares caused by data siloed in centralized repositories. Giving up that freedom for the convenience of being able to spin up an AWS instance flies in the face of everything blockchain stands for.
An Alternative Future
Not all projects utilizing or developing decentralized storage solutions are paying lip service to the concept, it should be noted. Many are vigorously pursuing decentralization for its own sake and are doggedly refusing to take the financially cheaper but reputationally more expensive route offered by the cloud.
The decentralized content delivery network, Fleek Network, for example, has gone to great trouble to ensure that the users spinning up nodes are natively building and hosting them. Fleek’s nodes are designed to be hosted on local machines, and its architecture is structured in such a manner as to eschew cloud storage altogether.
As Fleek’s whitepaper acknowledges, “Traditional approaches typically involve infrastructure-based delivery, wherein users request their content from third-party servers. These CDNs are traditionally permissioned, trust-based, and subject to the control of dominating parties who can censor or limit content. This is incompatible with making web3 applications decentralized and end-to-end.”
Unfortunately, many projects give short shrift to the ethos and implementation of decentralization. It’s widely known, for example, that NFT projects routinely rely on conventional web storage to host their images. While the token the image is associated with may be uncensorable and eternal, the critical media attached can disappear and go offline the moment the project fades and its cloud payments stop.
Check Yourself or Wreck Yourself
Decentralization is about disintermediation. It’s about removing power from large players and the concomitant abuses that such structures supports. Until blockchain projects multilaterally make this their methodology, users risk being misled over the inviolability and whereabouts of their data. It may be cheap, ever-present, and reliable. But centralized storage embodies everything that’s wrong with web2. If web3 projects can’t embrace decentralized alternatives, what hope is there of legacy players ever making the switch?
For tech-savvy users, checking whether a project is using IPFS or similar architecture is relatively simple. For everyone else, beware of centralized storage in decentralized clothing. It’s not the solution you’re seeking.