Scaleway Outage History

Scaleway's FR-PAR region experienced a complete power failure in datacenter operator rooms during planned maintenance, causing network connectivity loss and service disruptions across all FR-PAR availability zones and DC5 for 27.4 hours. The incident affected instances, databases, object storage, serverless functions, and Kubernetes services with various issues including connection timeouts, latency problems, and volume attachment delays. The issue was resolved after datacenter teams replaced faulty batteries on all electrical paths and successfully switched back from backup generators to main power.

Scaleway's Kafka clusters became unreachable via private network following an infrastructure update, preventing users from connecting to their Kafka services. The connectivity issue affected all Kafka cluster access through the private network for 2.5 hours. The problem was resolved and all Kafka services returned to normal operation.

Scaleway's API Gateway experienced major network connectivity issues that prevented customers from accessing APIs across all regions for 19.7 hours. The incident affected API availability and connectivity, causing service disruptions for users trying to interact with Scaleway's services. The issue was fully resolved after nearly 20 hours of investigation and remediation work.

Scaleway experienced a PostgreSQL LISTEN/NOTIFY bug in the fr-par-1 region that prevented Asahi and DevOS configuration/installation workflows on Mac-Mini servers from completing successfully. This caused Apple Silicon machines to temporarily enter error states and significantly increased delivery and installation times due to failed workflows requiring retries. The incident was resolved after 2.7 hours of service disruption.

Scaleway's Identity and Access Management service experienced a major outage from 10:55 AM to 11:01 AM CET due to an internal dependency issue. The incident prevented all management actions on IAM resources, caused automatic console logouts, blocked console logins, and made several dependent services partially unavailable including SSH keys, Kubernetes, MongoDB, and other control plane endpoints. A fix was deployed at 11:01 AM CET that immediately resolved the root cause and restored all affected services.

Scaleway's Generative API experienced instabilities for 2.5 hours, causing customers to encounter 403 and 500 errors when querying the service. The root cause was identified and a fix was implemented to resolve the issue.

Scaleway's Kapsule service in the IT-Milan region experienced issues with cluster upgrades and new cluster creations due to an internal dependency problem. The incident affected all users attempting to upgrade existing clusters or create new ones in that region for 10 minutes. The service team actively worked on resolving the dependency issue.

Scaleway experienced an IPv4 address assignment issue in the Paris (fr-par-2) region that prevented the creation of new instances requiring public IPv4 addresses. The incident affected Instances, Kapsule, LBaaS, and Billing API services for 2.6 hours before being resolved.

Scaleway's WebHosting platform pf-1005 in the fr-par region experienced connectivity issues that prevented customers from accessing their hosting services. The incident affected the Hosting component and lasted 22 minutes. The issue was successfully resolved by Scaleway's technical teams.

Scaleway's Cockpit service experienced elevated error rates affecting metrics and logs functionality in the nl-ams and pl-waw regions for 3.8 hours. Customers encountered difficulties accessing or retrieving monitoring data and log information during this period. The issue was investigated and resolved, with service returning to normal operations.