Complete switchover processing

Graceful Switchover consists of multiple steps on both the primary and the standby server. After each step a checkpoint is initiated on both servers. Only until both servers have reached the same checkpoint will processing continue to the next step. This ensures that both servers are always in a consistent state and rollback or roll forward is possible in case there are any issues. Backup copies of the control files, the redo logs and the database configuration files are made in case a rollback has to be initiated. The backup copies are in the same location as the original files and have a date stamp suffix in the form yyyymmddhh24mi. Redo logs and the archive logs are compressed before being transferred to the other server. 
When an error occurs on either the primary server or the standby server, processing will stop on both the primary and standby server.

The complete switchover process is shown below:

1. The process on the primary server to transition from a primary database to a standby database:

2. The process on the standby server to transition from a standby database to a primary database: