Database high availability best practices , for single-instance,RAC databases and data guard standby database.
- Always use spfile
- Multiplex control files
- Set control_file_record_keep_time long enough
- Multiplex production and standby redo log
- Log checkpoint to the alert
- Use auto tune checkpoints
- Enable archive mode and use flash recovery area
- Enable flashback database
- Enable block checking (db_block_checksum=true)
- Use automatic undo management
- Use locally managed tablespace
- Use segment space management auto
- Use re-sumable space allocation
- Use database resource manager
- Register all instances with remote Listeners
- Use Temporary tablespaces
Reference: Oracle database 10g : Real Application clusters (d17276gc10) p.11-16
