Refer to DM8 backup and recovery manual (version 1-1-190)
Scenario in which the active and standby clusters are restored to a single machine
Restore the archived log files of the primary database and the standby database to the stand-alone database.
Environmental description
OS: Linux
DB_VERSION: DM8 1-1-190-21.04.16-137954-ENT Pack4
The primary and standby clusters use the backup database archive to restore and recover
The first step is to back up in the main library
[dmdba@localhost bin]$ ./disql disql V8 user name: password: The server[LOCALHOST:5236]:The main library is open Login usage time : 5.839(ms) SQL> SQL> create table t as select * from dba_objects; Operation executed Elapsed time: 742.287(millisecond). Execution number:400. SQL> select count(*) from t; Line number COUNT(*) ---------- -------------------- 1 1365 Elapsed time: 4.000(millisecond). Execution number:402. ---Perform backup SQL> backup database full backupset '/home/dmdba/DB_BAK_FULL01'; Operation executed Elapsed time: 00:00:03.972. Execution number:404. ---Create two more tables after the backup is complete SQL> create table ttt as select * from dba_objects; Operation executed Elapsed time: 68.788(millisecond). Execution number:500. SQL> create table tt as select * from dba_objects; Operation executed Elapsed time: 89.548(millisecond). Execution number:507. SQL> select count(*) from ttt; Line number COUNT(*) ---------- -------------------- 1 1367 Elapsed time: 0.778(millisecond). Execution number:601. SQL> select count(*) from tt; Line number COUNT(*) ---------- -------------------- 1 1369 Elapsed time: 1.011(millisecond). Execution number:602. SQL> exit
The standby database switches online logs and executes checkpoint to generate archive logs.
Step 2: copy the backup files and backup archive logs to the destination database
[dmdba@localhost ~]$ scp -r DB_BAK_FULL01 192.168.15.50:/home/dmdba dmdba@192.168.15.50's password: DB_BAK_FULL01.bak 100% 27MB 32.8MB/s 00:00 DB_BAK_FULL01_1.bak 100% 11KB 5.3MB/s 00:00 DB_BAK_FULL01.meta 100% 81KB 23.5MB/s 00:00 [dmdba@localhost arch]$ scp *.log 192.168.15.50:/data/dmarch/stb_arch dmdba@192.168.15.50's password: STANDBY_ARCHIVE_0x137E5B0D[0]_2021-05-11_18-30-39.log 100% 25KB 7.8MB/s 00:00 STANDBY_ARCHIVE_0x137E5B0D[0]_2021-05-11_18-35-34.log 100% 128MB 57.5MB/s 00:02
The third step is to restore in the single machine destination database
It should be noted that:
By default, in the recovery phase, only the database DB to be recovered is collected when collecting archives_ MAGIC,PERMANENT_MAGIC consistent archive files, while permannet in the active and standby environment_ Magic is consistent. At this time, you can specify the DB of the archive log of the standby database_ Magic skips this limit if DB is not specified_ Magic will report an error.
(1) You can use DMRMAN to view the DB of the backup file of the primary database first_ Magic value:
RMAN> show backupset '/home/dmdba/DB_BAK_FULL01'; show backupset '/home/dmdba/DB_BAK_FULL01'; <backupset [DEVICE TYPE:DISK, BACKUP_PATH: /home/dmdba/DB_BAK_FULL01] info start ..........> <DB INFO> system path: /dm/dm134/data/DAMENG pmnt_magic: 197080140 src_db_magic: 327047949 db_magic: 327047949 ---Of backup files DB_MAGIC dsc node: 1 sys mode: 1 page check: 0 rlog encrypt: 0 external cipher[id/name]: 0/ external hash[id/name]: 0/ length in char: 0 use new hash: 1 page size: 16KB extent size: 16 case sensitive: 1 log page size: 512B unicode_flag/charset: 1 data version: 0x7000C sys version: V8 enable policy: 0 archive flag: 1 blank_pad_mode: 0 crc_check: TRUE page_enc_slice_size: 0 <META INFO> backupset sig: BA backupset version: 0x400A database name: DAMENG backup name: DB_FULL_20210511_182553_195320 backupset description: n_magic: 0x3F029C24 parent n_magic: 0xFFFFFFFF meta file size : 82432 compressed level: 0 encrypt type: 0 parallel num: 1 backup range: database mpp_timestamp: 1620728753 with_huge: FALSE backupset_type: NORMAL mpp_flag: FALSE backup level: online backup type: full without log: FALSE end_lsn: 49319 max_trxid: 48413 base begin_lsn: -1 base end_lsn: -1 base n_magic: 0xFFFFFFFF base name: base backupset: backup time: 2021-05-11 18:25:57 min exec ver: 0x08010101 min dct ver: 4 pkg size: 0x02000000 <EP INFO> EP[0]: begin_pkg_seq: 4077 begin_lsn: 49304 end_pkg_seq: 4086 end_lsn: 49319 <FILE INFO> backupset directory: /home/dmdba/DB_BAK_FULL01 backupset name: DB_BAK_FULL01 backup data file num: 4 backup piece num: 2 backup huge file num: 0 <backup_piece_list> $file_seq |$size(KB) |$pos_desc |$content_type 0 |27430 |DB_BAK_FULL01.bak |DATA 1 |11 |DB_BAK_FULL01_1.bak |LOG <data_file_list> $file_seq |$group_id |$group_name |$file_id |$file_path |$mirror_path |$file_len |$copy_num |$size_flag 1 |0 |SYSTEM |0 |/dm/dm134/data/DAMENG/SYSTEM.DBF | |45088768 |0 |0 2 |1 |ROLL |0 |/dm/dm134/data/DAMENG/ROLL.DBF | |134217728 |0 |0 3 |4 |MAIN |0 |/dm/dm134/data/DAMENG/MAIN.DBF | |134217728 |0 |0 <arch_file_list> $file_seq |$dsc_seq |$file_len |$begin_seqno |$begin_lsn |$end_seqno |$end_lsn 4 |0 |6144 |4077 |49304 |4086 |49319 <huge_file_list> $group_id |$schema_id|$table_id |$column_id|$file_id |$file_len |$path <backupset [DEVICE TYPE:DISK, BACKUP_PATH: /home/dmdba/DB_BAK_FULL01] info end .> show backupsets successfully. time used: 26.900(ms) RMAN>
(2) Use the dmrachk tool to view the DB of the backup archive log file_ Magic value:
[dmdba@localhost bin]$ ./dmrachk help rachk V8 format: dmrachk keyword=value routine: dmrachk arch_fil=/opt/a.log Mandatory parameters: Must specify arch_fil perhaps arch_path keyword explain(Default value) -------------------------------------------------------------------------------- ARCH_FIL Specify an archive CMP_FIL Specify an archive for comparison ARCH_PATH Specify an archive directory CMP_PATH Specify an archive directory for comparison TRUNC_FIL Specify a new file to store valid data extracted from the error archive. EXTRACT_FIL Specify a new file to store the data from the specified seq Extract data from a range of archive files ARCH_BAK_PATH Specify the backup file path to store duplicate files during archive repair CHECK Verification level(0, 1, 2, 3, 4) 0: File verification, 1: Package verification(seq & magic)，2: ptx Verification, 3: rrec check, 4: Clean up invalid archive files,default(0) BEGIN_LSN When specifying the archive directory, specify exclude less than that LSN Documents END_LSN When the specified archive directory is larger than the specified archive directory, exclude LSN Documents BEGIN_SEQ When specifying the archive directory, the extraction is greater than this SEQ Documents END_SEQ When specifying the archive directory, extract less than SEQ Documents BEFORE_LSN CHECK=4，Specify the archive directory and delete it LSN Previous documents. Default = 0, all LSN Delete all logs within the scope BEFORE_SEQ CHECK=4，Specify the archive directory and delete it SEQ Previous documents. Default = 0, all SEQ Delete all logs within the scope BEFORE_TIME CHECK=4，Specify the time point before deleting the file in the directory. Delete archives at all time points by default EXCLUDE_DB_MAGIC CHECK=4，Specify the archive directory and delete the specified DB_MAGIC Other documents. OUTPUT_FIL LOG You can specify output to a file DCR_INI appoint dmdcr.ini Path, if not specified, is not used DM_INI Target library dm.ini Path to get log encryption information AFTER_FREE appoint TRUNC_FIL Whether to check and extract FREE What follows. 0 no extraction; 1 extraction. Default 0 HELP display help information [dmdba@localhost bin]$ ./dmrachk ARCH_PATH=/data/dmarch/stb_arch rachk V8 the database pmnt_magic: 197080140 the database db_magic: 1125819454 /*******************************************************************/ Archive file/data/dmarch/stb_arch/STANDBY_ARCHIVE_0x137E5B0D[0]_2021-05-11_18-30-39.log detailed. version : 0x7006 status : INACTIVE enc_flag : FALSE n_rpkgs : 2 pmnt_magic : 197080140 db_magic : 1125819454 ---Of backup archive log files DB_MAGIC src_db_magic : 327047949 arch_lsn : 50279 arch_seq : 4102 clsn : 50510 next_seq : 4103 g_next_seq : 4103 file len : 25088 file free : 25088 dsc seqno : 0 dsc node : 1 create time : 2021-05-11 20:45:58.000000 close time : 2021-05-11 18:31:54.429738 /*******************************************************************/ /*******************************************************************/ Archive file/data/dmarch/stb_arch/STANDBY_ARCHIVE_0x137E5B0D[0]_2021-05-11_18-35-34.log detailed. version : 0x7006 status : ACTIVE enc_flag : FALSE n_rpkgs : 16 pmnt_magic : 197080140 db_magic : 1125819454 src_db_magic : 327047949 arch_lsn : 50511 arch_seq : 4104 clsn : 51779 next_seq : 4119 g_next_seq : 4119 file len : 134217728 file free : 168960 dsc seqno : 0 dsc node : 1 create time : 2021-05-11 18:31:54.429738 close time : 2021-05-11 18:35:34.056232 /*******************************************************************/ Summary (node)[0]): total: 2 Files correct: 2 Files error: 0 Files repeat: 0 Files the rachk tool running cost 2.969 ms
(3) Start restore
RMAN> restore database '/data/dmdata/TEST2/dm.ini' from backupset '/home/dmdba/DB_BAK_FULL01';
(4) Start recovery
If you do not specify the DB of the backup archive file_ Magic, using the default DB_ Magic (that is, DB_MAGIC of the primary database, DB_MAGIC in the backup file of the primary database, and src_db_magic in the archive log of the backup database) will report an error. As follows:
Specifies the DB of the archive_ Magic performs recovery:
RMAN> recover database '/data/dmdata/TEST2/dm.ini' with archivedir '/data/dmarch/stb_arch' use db_magic 1125819454;
Update DB_MAGIC:
RMAN> recover database '/data/dmdata/TEST2/dm.ini' update db_magic;
Step 4 start the database and OPEN to verify the data
After the main database backup is restored to a single machine, it can only be started to mount state by default. You need to change the database to normal mode and OPEN manually.
SQL> alter database normal; SQL> alter database open; SQL> select count(*) from t; SQL> select count(*) from ttt; SQL> select count(*) from tt;
At this point, the restore is complete.
epilogue
The scenario in which the primary and standby clusters use the backup archive for recovery is quite practical. After the primary archive is deleted, we can use the backup archive log to restore the data to the latest state.
When using the backup file of the primary database + the archive log of the standby database for restore and recovery, you need to specify the DB of the archive log file of the standby database during recovery_ Magic, if DB is not specified_ Magic will report an error. Because by default, only the database DB to be recovered is collected during archive collection in the RECOVER phase_ MAGIC,PERMANENT_MAGIC consistent archive files, while permannet in the active and standby environment_ Magic is consistent, so you need to specify DB_MAGIC to skip this limit.