Early on Oct 20, an Amazon Web Services (AWS) outage in the US-EAST-1 region cut connectivity to many cloud-dependent devices. Owners of Eight Sleep’s ‘Pod’ mattress covers lost remote control. Many reported overheating, while others found mattresses stuck in raised positions until cloud services returned.
What happened
The outage began around 03:00 ET on Oct 20, when AWS reported increased error rates and high latency in the US-EAST-1 region. Cloud-hosted controls for Eight Sleep’s devices became unreachable. Reports and social posts showed users were unable to change temperature or position settings while the backend was offline Dexerto.
How the devices were affected
Eight Sleep’s products rely on manufacturer servers to process commands, log biometrics, and manage water-cooled coils. When AWS failed, the app could not send updates to devices. Units generally remained at the last commanded setting, which left some beds heating and others failing to cool. Several owners posted images and short clips that showed beds locked at elevated temperatures or stuck in an inclined state.
Company response and user evidence
Eight Sleep posted updates on X (formerly Twitter). The CEO acknowledged the disruption and said the company was restoring features as AWS recovered. Users shared temperature logs, videos, and timestamps that matched the outage window. Monitoring services and social platforms recorded spikes in incident reports during the same period.
Technical context and design choices
Cloud-first device architectures centralize control and telemetry to vendor servers. This model reduces device complexity but increases reliance on network and provider resilience. Without an offline fallback or local manual override, a loss of cloud connectivity removes critical controls. In Eight Sleep’s case, firmware and design choices meant some functions defaulted to the last saved state, rather than a safe, manual fallback.
Impact and scale
Public reporting indicated issues across multiple US states. Dexerto compiled user reports, and several widely shared posts went viral. Eight Sleep did not publish an official affected-units tally. The broader AWS event also affected many apps and services, which increased visibility and user concern about cloud dependencies.
Operational timeline
- Oct 20, 03:00 ET — AWS reports increased error rates and latency in US-EAST-1.
- Oct 20 — Owners of Eight Sleep Pods report beds overheating or stuck in raised positions.
- Oct 20–21 — Social evidence and monitoring services show increased reports for related consumer services.
- Oct 21 — Eight Sleep posts updates confirming staged restoration as services returned.
Broader implications
This incident underscores the trade-offs in cloud-dependent consumer devices. Centralized backends enable feature updates and data collection. They also create single points of failure when providers experience outages. Similar vendor-supply chain discussions followed other incidents earlier in the year; for example, our coverage of the F5 Networks breach explains how infrastructure faults can cascade across customers.