Abstract: Techniques are provided for performing a failover for one or more streaming data applications executing in a cloud computing environment. An active application may actively communicate with a logical region of the cloud computing environment. A corresponding standby application may operate in standby mode with a different logical region. The active application and its standby application may each monitor the health of the logical region and the different logical region by polling respective health status applications. When it is determined that the logical region has failed and the different logical region is healthy, the active application may transition to standby mode and the standby application may transition to active mode to actively communicate with the different logical region. When it is determined that the logical region is healthy again, the applications may switch back to their initial modes.