The redo log files are written by the background process Log Writer(LGWR) with the information stored at the log buffer. If the log files are not available the process gets stuck, the transactional activity in the database is frozen and an error message is written to the alert.log
file (ORA-00313: open failed for members of log group ## of thread ##
), as shown in the next screenshot.
When a loss of redo log files occurs there are two different scenarios which depend on the status of the lost redo log. If the lost redo log had the status of inactive it is not critical. It is just enough to logically remove the group and rebuild it. On the other hand, if the status of the lost redo log file was CURRENT
, then the situation may get complicated as the CURRENT
status means the redo log was at that time written by the LGWR process, and the DB Writer(DBWR) process had not yet synchronized the datafiles by writing the dirty blocks from the DB buffer cache to the datafiles...