Slab Outage History

This incident has been resolved..

A recent update to part of the cache system introduced performance issues that caused the application to become unresponsive. The initial rollback did not immediately resolve the problem, as the issue was triggered only after the cache reached a specific threshold. A fix has since been implemented and deployed, and the application has been operating normally for the past several hours..

During a routine application update, a long-running database migration caused newly deployed servers to experience errors. Existing servers continued operating normally. The migration has been optimized, the new version successfully deployed, and all systems are now fully operational..

During a routine deployment, our application and APIs became unavailable due to the unavailability of a legacy open-source container image, which also prevented rollback. We updated our configuration to use the maintained version of the resource, and deployments are now proceeding normally. The application and APIs are fully accessible again. We are reviewing our dependency management processes to prevent similar issues in the future..

This incident has been resolved. We have finished our investigation into this issue. The root cause was a recent change we made which inadvertently caused some draft posts to return an error when loaded. This error resulted in increased memory usage, degrading performance and preventing some posts from loading or saving in a timely manner. In the process of resolving the first issue, we encountered a separate, unrelated caching problem which required additional time to resolve. Both issues were ultimately resolved by restoring a previous version of the application. We have been monitoring our error rates and memory usage, and they have remained stable since the incident's resolution..

All application functionality have been observed to be operating normally.