Server Repave Notification

The application runs as part of a distributed cloud computing infrastructure. This infrastructure allows for automated switch-over in the event of failure, and for repaving switch-over for regular hardware rotation and patching. Periodically (typically every 14 days), the application will therefore repave to run on a different server instance. When this occurs, any open sessions are transferred to another running pod. The time taken to transfer these sessions is dependent upon the number of active users on the pod at that particular instance. In case of high usability, it is possible that a timeout on the browser could be experienced.

In order to handle the repaving in a way that the user gets the opportunity to reconnect to the new server instance prior to the point of switch-over, a notification is issued that the server has a limited lifetime remaining, so the user must save any changes and log out then log on again, otherwise risk being disconnected and losing any unsaved changes. When the user logs back on, the session will automatically connect to the new server instance and clear the notification.

When the system detects that repaving is imminent, the notification is shown as pop-out toast message.