Abstract: System and methods for temporary data management in shared disk cluster configurations is described. In one embodiment, a method for managing temporary data storage comprises: creating a global temporary database accessible to all nodes of the cluster on shared storage; creating a local temporary database accessible to only a single node (owner node) of the cluster; providing failure recovery for the global temporary database without providing failure recovery for the local temporary database, so that changes to the global temporary database are transactionally recovered upon failure of a node; binding an application or database login to the local temporary database on the owner node for providing the application with local temporary storage when connected to the owner node; and storing temporary data used by the application or database login in the local temporary database without requiring use of write ahead logging for transactional recovery of the temporary data.