I’m completely stumped as to why the WebRTC softphone client fails to initialize for agents working the early morning shift in America/Chicago. The issue manifests specifically when agents attempt to log in before the main WFM schedule adherence checks run at 06:00 CT. The browser console throws a WebSocket connection to 'wss://api.mypurecloud.com/websocket' failed error, followed immediately by a 503 Service Unavailable response on the /api/v2/users/me/phone endpoint. This seems to correlate with the high volume of concurrent schedule publishing events occurring simultaneously.
We are using the latest Genesys Cloud SDK version 1.24.0. The problem does not occur during the mid-day shift, suggesting a potential resource contention issue between the WFM publishing engine and the signaling service. The agents’ desktop environments are standardized Chrome instances with no conflicting extensions. Restarting the browser or waiting 15 minutes resolves the issue, but this downtime impacts our schedule adherence metrics significantly. Has anyone seen this specific handshake failure tied to WFM API load?