Format is used to define the path and naming rules of backup files.
- Path: If it is definedDb_recovery_file_dest, The backup file is generated in this directory! If not defined, % ORACLE_HOME % \ database [DBS] will be generated in the default system location
- Format: the default format is include% UOMF string
- Recommended format definition :{% D | % I}-{dB | TB | DF | ar | CF | SP}-% T-% s-% P-% C-% T. bkp-- Database name-Basic Object Type-year, month, and day-Backup set serial number-Backup part copy count-timestamp. bkp
You can specify the format in many places and runtime environments. The system always defines the override with a higher priority.
- Bakupspec
- Backup
- Allocate Channel
- Configure
Filesystem characters allowed by any OSYes, but it should passReplace variableImplemented file nameUniqueness!
FormatReplace variable, Case sensitive!
- % d -- db_name of the database
- % N -- db_name of the database with 8-bit length. Fill the remaining part with "X".
- % N -- Name of the database tablespace
- % I -- dbid of the database
- % t -- yyyymmdd = % Y % m % d
- % t -- timestamp with 9 Characters
- % s -- backup set number
- % P -- backup part number
- % C -- the serial number of multiple copies of the backup slice
- % E -- the serial number of archived redo file. it can only be used on archived redo.
- % f -- datafile filenmuber, it can only be used to back up datafile and tablespace; otherwise it does not make sense
- % f -- composite format = C-IIIIIIIIII-YYYYMMDD-QQ, where iiiiiiiiii is dbid, yyyymmdd is year month day, QQ is the serial number (00-ff) of multiple copies of the hex backup slice ). TMD, Oracle prompts error!
- % u -- 8 is a unique letter string
- % u -- composite format ==
- Backupset: % u _ % P _ % C
- Copy of datafile: Data-D-% d_id-% I _ts-% n_fno-% F _ % u
- Copy of archived log: arch-D _ % d-ID-% I _s-% e_t-% h_a-% A _ % u
- Copy of controlfile: CF-D _ % d-ID-% I _ % u
