Chatting with a friend who is doing Web3 development, he complained with a frown about being stuck in a dead end with decentralized social dApps due to storage issues. He tried various methods but to no avail: the cost of storing on L1 chains is ridiculously high, and large files like short videos and audio simply can't be handled; using IPFS solves the addressing problem, but there is no guarantee of data persistence, and obscure content can disappear just like that. Finding third-party storage services incurs extra costs and compromises the original intention of decentralization, leaving him in a real dilemma.
I immediately told him to quickly try Walrus, a storage solution developed by Mysten Labs, specifically designed to address this pain point, aimed at solving Web3 storage challenges. Other storage solutions are stuck on consensus, tying storage and node validation together, forcing all nodes to carry full data, which is both slow and expensive. On the other hand, Walrus completely separates the coordination management and storage nodes of Sui, allowing validation nodes to avoid the data burden, with dedicated storage nodes handling large unstructured files like video and audio, maximizing efficiency.
More importantly, it does not rely on mindless redundancy for security but uses erasure coding and a self-developed Red Stuff algorithm to achieve data recovery through mathematical probability, significantly reducing storage costs by several orders of magnitude. The short videos and audio in your dApp, even the AI model weights you might need to add later, can be stored cheaply and natively on a decentralized network, eliminating the need to secretly store large files on AWS while pretending to be decentralized.
After hearing this, my friend’s eyes lit up immediately, stating that this is the kind of solid infrastructure that solves real problems, not just a flashy concept. I urged him to quickly test it out on the testnet to see the actual upload and retrieval speeds, as well as the data recovery capabilities during node fluctuations. This will definitely resolve the storage issue that has been a bottleneck. After all, for those working on Web3 applications, a solid storage solution like Walrus is truly a necessity.
