1、Checkpoint Process (CKPT)：
The checkpoint process (CKPT) is responsible for signaling DBWn at checkpoints and updating the datafiles and control files of the database to indicate the most recent checkpoint.
Every three seconds, CKPT records the RBA from the oldest entry in the checkpoint queue in the control file. This RBA represents the point in the redo log at which instance recovery is to begin after an instance failure. It can do this because all of the data blocks represented in prior redo records are guaranteed to have been written to disk by DBWn.
Only in the event of a log switch, does CKPT also write this information to the headers of the datafiles. To improve performance not all of the headers are written at once, but rather they are updated in groups in what is referred to as a “lazy write.” Log switches do not force all dirty buffers to be written to disk.
2、Control File
The control file is a binary file that describes the structure of the database. It must be available for writing by the Oracle server whenever the database is mounted or open. Its default name is operating system-dependent. Without this file, the database cannot be mounted and recovery or re-creation of the control file will be required. The recommended configuration is a minimum of two control files on different disks to minimize the impact of a loss of one control file.
Control File Contents
• Database name
• Time stamp of database creation
• Synchronization information (checkpoint and log sequence information) needed for recovery
• Names and locations of datafiles and redo log files
• Archiving mode of the database
• Current log sequence number
• Recovery Manager backup meta data
3、Archive Background Process
The ARCn process is an optional process. When enabled, it archives the redo log files to designated storage areas. This process has a great significance in backup, restoration, and recovery of a database set to ARCHIVELOG mode, where databases are operational 24 hours a day and 7 days a week.
The ARCn process initiates when a log switch occurs and copies one member of the last (unarchived) redo log group to at least one of the destinations specified by one or more initialization parameters.