Web3 use cases for storage flimsy for now

Future unwritten

Storage administrators looking to dig into Web3 have their work cut out for them. As of right now, the term itself is still unclear, let alone its value to the enterprise.

Martha Bennett

Avivah Litan

"There's no agreed-upon definition of Web3," said Martha Bennett, principal analyst at Forrester Research. "It's been appropriated by the blockchain ecosystem. It's largely a designation label to say the future will be blockchain and crypto-economics."

Other analysts, however, say Web3 has already been defined and includes a stack of technologies that are joined at the hip to decentralization not just of storage, but of other information commonly tied into existing database concepts.

"In Web3, users control their own algorithm," said Avivah Litan, an analyst at Gartner, citing the definition outlined by the Web3 Foundation, a consortium of developers that support blockchain projects, education and building compatibility protocols. "It's not subject to interpretation, frankly."

A user-controlled algorithm, according to Litan and the Web3 Foundation, will include defining exactly what content is displayed and collected about the user, usage of the blockchain cryptocurrencies for payment and high encryption levels.

None of that definition, however, intrinsically meshes with enterprise storage concerns and needs.

"Blockchain is just metadata," Litan said. "You never want to use it as a database. You can trust the metadata, but you never want to store the data on the blockchain. You always store the data off chain."

Marc Staimer, president of Dragon Slayer Consulting, believes storage will remain a commodity underlying technologies such as blockchain. The blockchain may assure your personal ownership of a JPEG, but the storage technology to back up the image isn't a Web3 technology, according to Staimer.

Marc Staimer

Staimer felt the current iteration of Web3 is little more than marketing glitz on existing services such as targeted advertising, web interactivity and digital purchases.

"It's not a change from what we're doing right now; it's an articulation of what we're doing," Staimer said. "Putting a label on it is kind of silly. Storage vendors are all trying to differentiate their product because all storage is commodity."

Paul Nashawaty

That need for storage vendors to differentiate in the market, primarily through advertising, will create market confusion, according to Paul Nashawaty, a senior analyst at Enterprise Strategy Group, a division of TechTarget. Additionally, potential enterprise customers will find a lack of other users and product information until adoption picks up, so investing in Web3 technology now is risky.

"There's definitely blockchain services available, but it goes back to the maturity of the organization," Nashawaty said. "You have to be able to take those services and implement them. They're not an end-to-end service."

Distributed services

Aggregators and advocates of distributed storage networks believe there is an enterprise future for the technology, despite most offering primarily consumer-grade services.

Filebase, one such vendor, sells S3-compatible object storage services built on top of blockchain storage networks including Sia, Skynet and Storj. Filebase acts as a front end and aggregator for the three networks, allowing users to select their specific network choices and storage locations through a console akin to creating an Amazon S3 bucket without needing to set up specific network connections or configurations.

Filebase users, all of which are currently consumers, can store up to 5 GBs of data free of charge. A paid subscription for 1 TB of storage and 1 TB of bandwidth is available for $5.99 per month, with additional storage available at $0.0059 per GB of storage and outgoing bandwidth.

Storage vendors are all trying to differentiate their product because all storage is commodity.

"Web3 allows us to build an exabyte-level cloud without having to build additional data centers," said Zac Cohen, co-founder and COO at Filebase. "We are blessed in the fact that the more networks we add to our platform, the more capacity we have access to."

Sia, Skynet and Storj, however, transact in cryptocurrency. Connecting to the networks requires a user to mine for and maintain a balance of cryptocurrency to pay for and maintain their files across the network, a process that taxes both power consumption and hardware and can be expensive.

Once cryptocurrency is acquired, the price of a given coin can fluctuate wildly depending on factors that are outside of the users' control, such as market demand or law enforcement investigations.

Filebase handles the cryptocurrency mining and balance maintenance for its customers, requiring just a customer email address to access the distributed networks.

"We certainly believe most customers don't care about this [setup]," Cohen said. "They just want cheap, fast and performant storage. "

Although Filebase's offerings currently target individual storage needs, the company has begun to branch out with additional services applicable to individual developers and enterprise users.

Filebase recently integrated with Flexify.IO, a multi-cloud storage virtualization and migration software company. The partnership enables users to choose Sia, Skynet and Storj in addition to other cloud storage providers, such as Wasabi and Backblaze, as well as the three hyperscalers, AWS, Azure and Google Cloud Platform (GCP).

Flexify.IO uses a cloud-agnostic API to connect disparate storage and cloud application services together, in pursuit of enabling users to connect to the cheapest services.

"It's about data access and data availability," Cohen said. "Your data is no longer siloed to a single data center or region of the internet."

Analysts, however, remain unmoved by the distributed storage pitch from companies like Filebase for enterprise users.

Forrester's Bennett said the inherent financial risk of cryptocurrencies, the difficulty in connecting to blockchain networks and the unpredictability of the service leaves most enterprise users cold.

"Enterprises do not want to deal with cryptocurrency," she said. "They want secure storage, redundant storage. What they don't need is another currency on the books that fluctuates wildly. You need SLAs, you need clear contracts, you need choice and control over where your data sits. A loss of that isn't covered by those systems."

Indeed, many distributed networks use sharding to break up larger files into smaller, more easily stored pieces across the network. These pieces can be encrypted and stored across user storage devices connected to the network, creating redundant files for recovery and ensuring no individual has complete access to a file.  

This implementation of sharding data for distributed storage can result in unacceptable access speeds and a level of unpredictability for enterprise uses, Bennett said.

"All you need is one provider down," she said.

Blockchain for applications

Larger enterprise vendors have kept blockchain technology at arm's length or have outright canned these products.

Around 2018, cloud hyperscalers and legacy enterprise vendors made a push to get into the distributed infrastructure game and announced either new products or partnerships to provide blockchains and distributed ledgers.

IBM, SAP and Oracle maintain blockchain products, along with AWS and GCP, ranging from managed services to IT starter kits for building off the open source Hyperledger Fabric framework, a permissioned blockchain infrastructure to create blockchain-based software and applications.

Most services remain available, although Microsoft Azure retired its Azure Blockchain Service in September 2021. But none of these enterprise applications focus on distributed storage or a blockchain implementation of storage.

"Blockchain is never going to be a storage system," Gartner's Litan said. "It just needs a persistent storage system."

Litan said many existing enterprises operate counter to the Web3 model. Storage administrators want to be able to track information through centralized databases rather than managing decentralized components, which they discovered when they attempted blockchain projects.

"Companies jumped on the bandwagon because of FOMO [fear of missing out]," she said. "The reason people walked away from it was they never implemented blockchain to begin with."

Some large enterprises still see a future in distributed ledgers. Financial companies are using Hyperledger Fabric, seeing value in an immutable chain of ownership and data changes. Litan said the adoption of cryptocurrency and maturation of existing blockchains will more likely draw in companies into a Web3 future.

"As soon as the enterprises can start using the existing public blockchain, then it'll be a different story," she said.

Staimer echoed Litan's sentiment that vendors haven't successfully pitched how blockchains can effectively replace existing services, and, in many cases, enterprises would have to give up speed or performance of existing technology for blockchain.

"By implementing blockchain [an enterprise's] interaction is slowed down, and they leave," he said.

Plus, the use cases for storage are more hype than reality, at this point.

"I don't see the application outside of finance. Maybe personably identifiable information, but that has nothing to do with Web3," Staimer said. "That's like saying, 'We have SSDs; that's Web 3.0.' I look at this and I go, 'I'm kind of stumped.'"

Tim McCarthy is a journalist living in the North Shore of Massachusetts. He covers cloud and data storage news.