After restoring the InTrust server, you can roll back this switching session (switch sites and jobs back to the server initially responsible for their processing). Rollback is a manual process, performed as follows:
Rollback history is stored in the configuration database. Once the active (failed) InTrust server has been restored and all job/sites failed back, the only communication between remote agents and the standby server is the keep-alive packet.
After the failure has been resolved, you might have to manually change the indexing servers in the properties of your repositories. These settings are not rolled back by the wizard.