- AWS has been unavailable for three hours
- Fortunately, this did not affect Ethereum
- But it could because of the excessive centralization of the network
- Most providers use cloud services, including AWS
Yesterday, June 13, the Amazon Web Service cloud service went down for several hours. As a result, a blackout occurred that affected dozens of sites, including major news portals (AP and Boston Globe) and food ordering services (Burger King, Taco Bell).
The exact cause of the failure is still unknown. At Amazon declarethat the issue has been resolved and all systems are operating normally.
The bug probably originated in the AWS Lambda engine. It is a service for deploying applications without the need for direct user control.
So how is Ethereum related to this?
According to ethernodesthe performance of more than 64% of network nodes directly depends on Amazon hosting providers:
The crash lasted three hours. And, quite naturally, the community had fears that Ethereum would “fall”, but it did not.
But on opinion Evan Van Ness, such a scenario is likely if the failure occurred in Europe. The reason for this lies in the staking provider Lido, which operates on the basis of cloud servers.
Ethereum has been repeatedly criticized for being too centralized. Including, because of the “dependence” on the Infura infrastructure, which provides network nodes to providers. Most of them, Lido in particular, use AWS. Earlier we covered a new article by Vitalik Buterin. In it, he talked about three important “transitions”, without which the further development of the ecosystem is impossible.