Coinbase to review speed-vs-resilience after AWS outage

An AWS cooling failure on May 7 halted Coinbase trading, blocked some account access and delayed balance updates; CEO Brian Armstrong called the outage “unacceptable.”

An AWS data center cooling failure in a single Availability Zone late on May 7 caused an outage that knocked several Coinbase exchange services offline, prevented some customers from placing trades and delayed display of account balances. The disruption affected spot trading as well as Prime, international and derivatives venues.

Coinbase halted trading when exchange systems could not operate safely during the infrastructure problem. Internal messaging slowed, which delayed account updates while teams worked to restore systems and reconcile data.

Engineering teams moved impacted workloads away from the troubled area of the data center, restored the systems needed to process orders and let lagging customer information catch up. Markets reopened in stages: cancel-only mode, followed by product checks and an auction phase, before full trading resumed on Coinbase Exchange.

Brian Armstrong wrote on X that the outage was “never acceptable.” He noted most Coinbase systems are designed to tolerate a failure in one Availability Zone but that the centralized exchange did not perform that way during this event. He said Coinbase will review how it balances exchange speed, customer co-location and recovery time after infrastructure failures, adding that the duration of an outage should be reduced considerably when an AZ move is needed.

Rob Witoff, head of engineering, posted an account of the recovery that emphasized customer impact. He described losing account access, even briefly, as unacceptable and explained that sequencing the recovery rather than restarting all systems at once was intended to preserve market safety while bringing customer-facing services back online.

Coinbase described the tradeoffs involved: many high-frequency and institutional traders colocate servers near an exchange’s matching engine to cut milliseconds from order execution. Redundancies that survive data center or zone failures typically add routing or synchronization steps that can increase latency.

Coinbase has not released a detailed timeline for infrastructure changes. The company said it will review resilience measures and recovery processes to reduce the likelihood and shorten the duration of similar outages in the future.

The material on GNcrypto is intended solely for informational use and must not be regarded as financial advice. We make every effort to keep the content accurate and current, but we cannot warrant its precision, completeness, or reliability. GNcrypto does not take responsibility for any mistakes, omissions, or financial losses resulting from reliance on this information. Any actions you take based on this content are done at your own risk. Always conduct independent research and seek guidance from a qualified specialist. For further details, please review our Terms, Privacy Policy and Disclaimers.

Articles by this author