Production app and its embedded booking calendars have been returning 502 since 09:35 UTC today (7 September). Please investigate a stuck container lifecycle/provisioning state urgently.

A new Docker deployment built successfully and pushed its image, then remained at Creating containers without application logs. The old process received SIGTERM. We aborted the new deployment, restarted the prior stable version and redeployed that exact old code/configuration. The rollback also failed before starting. SSH reports container status stopping, although the dashboard still called the old deployment Active. HTTP checks confirm it was not serving.

The separate MySQL service is Online and both persistent volumes are Ready and well below capacity. No database or volume migration or region change was requested. We have now explicitly stopped the stuck old deployment and launched a fresh redeploy of the exact stable version. Please check the attached service context and recover its container without deleting, detaching or resetting persistent volumes or database. There are no customer data or credentials in this report. Detailed deployment IDs can be supplied privately to Railway staff.