Original author: ChandlerZ, Foresight News
On April 15, Binance announced that the AWS data center network was temporarily interrupted and some of the platforms services had problems. Some orders were still successful, but some orders failed. All services are gradually being restored, and the withdrawal function has also been restored. During the full system recovery, some services may be delayed.
Binance is not the only one affected. Platforms including DeBank and Kucoin have also issued announcements stating that due to the AWS service outage, all DeBank services are temporarily unavailable. Their teams are continuously monitoring the situation and will restore access as soon as possible.
Amazon Web Services (AWS) is a comprehensive cloud computing platform that provides a wide range of on-demand services, including computing power, storage, databases, and other functions. According to the outage.now website, the platform has been down since 16:14 and has gradually returned to normal.
Although this incident did not trigger widespread on-chain asset risks or market fluctuations, it once again revealed an old problem: the crypto world claims to be decentralized, but at the most critical infrastructure level, it is still mired in the quagmire of centralized dependence.
AWS is a comprehensive cloud computing platform that provides a variety of services including computing power, storage, database and network transmission. Its platform hosts a large number of core Internet applications. In the Web3 ecosystem, many exchanges, wallet services, dApp front-end and back-end, oracle nodes, on-chain indexing services, etc. are built on the AWS architecture. Therefore, every AWS outage is not just a problem at the cloud service level, but a substantial impact on the underlying logic of the entire crypto industry.
Judging from this incident, whether it is centralized exchanges such as Binance and KuCoin, or Web3 wallets and asset tracking platforms such as DeBank and Zapper, they have all been affected to varying degrees after the AWS failure. In particular, the DeBank platform directly announced that all services are temporarily unavailable, indicating that even lightweight Web3 tools may be suspended due to the downtime of centralized cloud platforms.
This is not the first time the accident
Looking back, this is not the first time such an incident has occurred. In December 2021, a large-scale outage in the AWS US East (us-east-1) region became the most serious accident in recent years, affecting mainstream trading platforms such as Coinbase, Binance.US, Kraken, and dYdX. At that time, Coinbase users encountered API connection failures, transaction delays, and asset reading errors. dYdX even directly announced that the platform would suspend services. Although as a DEX, its contracts run on the chain, its front-end and data interface still rely on AWS services.
MetaMask was also affected at the time. Its default RPC service provider Infura, which connects to the Ethereum network, uses AWS for hosting, which caused some users to be unable to obtain on-chain data and transaction broadcasts to fail. Similarly, aggregators such as CoinMarketCap and Coingecko, which rely on AWS for data processing capabilities, also reported access delays and data update interruptions.
Another typical incident occurred in 2020, when AWSs Kinesis Data Streams service was interrupted, affecting a large number of platforms that rely on real-time data stream processing. Coinbase was affected again and its service was degraded. CoinGecko also reported problems with market data delays. This incident shows that the crypto fields dependence on AWS is no longer limited to web front-ends or basic hosting, but has penetrated deeper into real-time market data processing and the internal logic of trading systems.
The common point of these incidents is that when AWS has problems, not only the centralized trading platform services crash, but also the use of some decentralized protocols is seriously disrupted. This phenomenon of stable on-chain protocols and crashed off-chain interfaces is also one of the most tense contradictions in the current process of decentralization of Web3 infrastructure.
The challenges remain huge
Although the nodes of public chains such as Ethereum and Bitcoin are distributed around the world and will not be paralyzed due to service interruption in one place, a large number of Web3 projects still rely on AWS or other leading cloud service providers for user access portals, data synchronization paths, and front-end interfaces. In particular, some emerging projects often have their nodes deployed in the same area. Once the cloud service fails, the access capabilities of the entire network will be severely damaged.
As the demand for cloud resources from AI, Web3, and data services rapidly increases, the global cloud service market is showing a highly concentrated pattern. According to Statista data, by the end of 2024, AWS will have a market share of 31%, while Microsoft Azure and Google Cloud will rank second and third. This reality of high dependence on a few vendors makes single point failures in infrastructure not only a problem for one company, but also a systemic risk for the entire ecosystem.
Some developers and project owners in the industry have begun to explore alternative paths such as multi-cloud disaster recovery, IPFS front-end hosting, and Rollup autonomous node deployment to reduce dependence on Web2 infrastructure. However, these solutions often have practical challenges such as high development barriers, rising costs, and complex user experience, resulting in very few projects being truly implemented.
This AWS failure may not cause direct economic losses, but it has sounded another important alarm. Between the ideal goal of decentralization and the actual operating mechanism, Web3 still needs to complete a deeper infrastructure reconstruction.
