A hardware malfunction at an unnamed cloud provider is disrupting access to Ethereum, while the network’s consensus layer continues to operate normally. Between 35% and 60% of execution-layer nodes are hosted by cloud providers, including about 35% on AWS, exposing users to outages when centralized infrastructure fails. RPC providers reported minor performance issues in mid-November 2026, but no widespread Ethereum disruption was confirmed during the period examined. In October 2025, an AWS DynamoDB outage disrupted several Ethereum services for several hours without affecting consensus operations. The provider’s identity, the incident’s duration and the affected services remain unclear, while self-hosted nodes or access to multiple RPC endpoints could reduce reliance on a single provider.
Source: Read the original article
