Oracle Database-level drop redolog is not dangerous. Some DBAs are afraid to delete the redo log in the production database, because they are worried that deleting the redolog group in the ACTIVE and ACITVE status will cause a database accident. In fact, if the redo logs in the CURRENT and ACTIVE statuses are deleted from the database, they cannot be deleted. Therefore, you can safely delete the redo logs from the database. However, if you delete (rm) directly at the operating system layer of the Unix/Linux platform, ORACLE cannot perform the check, if the redo log in the CURRENT or ACTIVE State is directly deleted at the operating system level, it will lead to an accident and be very careful. Perform a REDOLOG test at the ORACLE database level with the deletion status of CURRENT and ACTIVE: 1. view the redo log status.
SQL>select group#,thread#,archived,status from v$log;GROUP#THREAD#ARCHIVEDSTATUS11NOINACTIVE21NOACTIVE31NOCURRENT41NOACTIVE
The above GROUP 3 status is CURRENT and GROUP2/GROUP4 is active. Now we will perform the delete test on the redolog group in the CURRENT status and ACTIVE status respectively. 2. Delete the group in the CURRENT status #3.
SQL> alter database drop logfile group3;alter database drop logfile group 3*ERROR at line 1:ORA-01623: log 3 is current log forinstance litest (thread 1) - cannot dropORA-00312: online log 3 thread 1:'/dba/oracle/oradata/litest/redo03.log'
3. delete an ACTIVE group #2
SQL> alter database drop logfile group2;alter database drop logfile group 2*ERROR at line 1:ORA-01624: log 2 needed for crash recoveryof instance litest (thread 1)ORA-00312: online log 2 thread 1:'/dba/oracle/oradata/litest/redo02.log'
4. Summary (1) when the redolog group in the CURRENT State is deleted, the system will prompt that it is the CURRENT log and cannot be deleted (2) when the redolog group in the ACTIVE state is deleted, this log is required for instance recovery and cannot be deleted.