Scaleway Outage History

A maintenance error caused all IP addresses in Scaleway's pl-waw-1 and pl-waw-2 zones to be released from the IPAM system, affecting Instance services. Users experienced internal errors when booking IPs, missing IPs from the IPAM API, and missing reverse DNS functionality. The issue was resolved after 17.4 hours with a fix implementation and monitoring period.

Scaleway's Public Gateway service in the nl-ams-1 region experienced issues with creating and updating gateways starting at 01:00 on February 28th. The incident affected gateway provisioning and modification operations in the Amsterdam region for 24.5 hours before being resolved on March 3rd.

Multiple databases in Scaleway's NL-AMS-1 region became stuck in intermediate states due to a Block Storage incident that began at 22:27 CET. The issue affected the Database-as-a-Service (DBaaS) component, preventing normal database operations. All databases were restored to healthy status by 12:10 CET, resolving the 1.9-hour outage.

A control-plane fault in Scaleway's nl-ams-1 block storage cluster caused multiple disks to be excluded simultaneously, resulting in progressive freezing of read/write operations for up to 31% of the cluster and affecting Instances, Kapsule, and DBaaS services. The storage team resolved the issue by reintegrating the excluded disks to restore cluster health, followed by product teams clearing remaining stuck instances and databases. No data loss was detected during the 5.6-hour incident.

Scaleway's Cockpit service in the fr-par region experienced errors when querying Logs datasources starting February 23, impacting users' ability to retrieve and analyze log data. The incident affected the fr-par-1 component and was resolved on March 6 after approximately 162.6 hours of disruption.

The API Gateway service in Scaleway's pl-waw region experienced connectivity issues with api.scaleway.com, preventing users from managing and accessing services through the API. The outage was caused by a misbehaving transit provider, which Scaleway removed from their network paths. The incident lasted approximately 14 days before being fully resolved.

Scaleway's transactional email service experienced delivery delays and retries when sending emails to Microsoft domains (Outlook, MSN, Hotmail) due to temporary rate limiting caused by IP reputation issues. The incident lasted over 6 weeks, affecting email delivery to Microsoft email services. The issue was resolved and email delivery returned to normal operation.

Scaleway performed a scheduled upgrade of a main backbone router at Paris DC3 on March 10th from 08:00 to 18:00 CET. The maintenance potentially affected services in Paris (fr-par-1, fr-par-2) and Amsterdam (nl-ams-1) regions despite the redundant router setup. The upgrade was completed as a minor maintenance event lasting 17.7 hours total.

A power outlet failure in Scaleway's nl-ams-1 datacenter caused 12 Elastic Metal and Dedibox servers to become unavailable. Technicians worked to restore power, with 11 servers recovered within approximately 2 hours and the final server restored after 14.3 hours. The incident was fully resolved with all affected servers back online.

A Dedibox switch in rack 6.12.53 at Scaleway's Amsterdam datacenter became unreachable due to a suspected power supply problem, making all connected Dedibox servers inaccessible for 135.4 hours. The issue was resolved by replacing the faulty switch and pushing the configuration, with customers advised to perform a DHCP request to restore access due to security features.