...
Each CAS server in the cluster has a shadow object representing the TicketRegistry of each of the other nodes. In normal operation , that object contains no ticket objects. There is no need to extract objects from the other node's files until a failure occurs and a request for one of those tickets arrives. Then Cushy restores the tickets from the file into memory and processes requests on behalf of the failed the CAS nodes exchange checkpoint and incremental files but they do not restore objects from those files to memory. This is called "Tickets On Request". The first time a request arrives for a ticket owned by another node, the getTicket request restores tickets into memory from the files for that node.
However, every new ticket Cushy creates belongs to the node that created it. During a node failure, the new Service Tickets or Proxy Granting Tickets created for users logged into the failed node are created by and belong to the backup node. They each get a ticket ID that has the suffix of the backup node. They live forever in the Ticket Registry of the backup node. They just happen to be associated with and point to a TGT in the shadown shadow registry on the backup node associated with the failed login node.
...