On May 15th, the Stellar network halted for 67 minutes due to consensus failure, but the ledger state remained safe and consistent. The outage revealed the need for better validator monitoring, faster restart procedures, and improved community standards around validator maintenance and configuration.
The Stellar network experienced a 67-minute outage on May 15th when it could not reach consensus, resulting in no ledgers being closed or transactions processed. However, the network's design prioritized consistency over liveness, preventing any data loss or forks. The halt was triggered when Keybase took down their validator for maintenance while other validators were already misconfigured or down. The Stellar Consensus Protocol functioned as intended by halting rather than risking an inconsistent state. The incident revealed that despite recent decentralization efforts, new validators lacked sufficient operational standards. The SDF outlined improvements including better monitoring and alerting systems, quicker validator restart procedures, improved onboarding for new validators, and better operator coordination around maintenance schedules. Within an hour of the halt, key validators coordinated a configuration change to restore network functionality.
