Major incident

Codeweavers: Resolved: Overnight Service Disruption

Current status of Codeweavers: Operational — live Codeweavers status

Timeline

1. Aug 26, 2026, 22:30 UTC
   resolved
   We experienced a period of service disruption between 23:55 BST on 26 August 2026 and 06:19 BST on 27 August 2026. The issue has now been resolved and services are operating normally. We are conducting a full review of the incident and will publish a Root Cause Analysis (RCA) once this has been completed. We apologise for any inconvenience caused.
2. Sep 7, 2026, 07:57 UTC
   postmortem
   ## Incident Summary On the evening of 26 August 2026, we experienced an issue within our database platform that resulted in customers being unable to access Codeweavers applications, plugins and APIs. The incident developed gradually as database resources became exhausted, ultimately causing the platform to become unavailable. Service was fully restored at 06:19 on 27 August 2026. Our investigation identified that a database request became stuck and prevented other requests from completing successfully. As additional requests accumulated, available database connections were eventually exhausted, preventing the platform from processing further activity. Once identified, engineering teams removed the blocked processes, restored database connectivity and completed recovery activities to return all services to normal operation. ## Customer Impact Between approximately 23:25 on 26 August and 06:19 on 27 August, customers experienced an interruption to platform services. During this period, access to applications, plugins, APIs and live platform data was unavailable. ## Timeline * **20:52** - Initial database transaction became trapped and started blocking subsequent activity. * **23:25** - Available database connections became exhausted, resulting in platform-wide service disruption. * **05:04** - Incident formally declared and engineering teams engaged. * **05:29** - Blocked database sessions were identified and removed, allowing recovery efforts to begin. * **06:16** - Platform recovery activities completed and services prepared for restoration. * **06:19** - Full service restored and monitoring confirmed normal operation. ## Corrective Actions A software change has been implemented to ensure database resources are correctly released when requests are cancelled, preventing the accumulation of inactive connections that contributed to this incident. In addition, we are reviewing our out-of-hours alerting and escalation processes to improve response times and reduce the duration of any future incidents.

Updates as published by Codeweavers (original).