https:\/\/blog.capdata.fr\/index.php\/aws-backup-restore-sql-server-rds-iaas-ec2-et-vice-versa\/<\/a><\/p>\n <\/p>\n
Quelques recommandations<\/h3>\n
AWS RDS offre la possibilit\u00e9 de faire des backup RMAN, en revanche, il ne sera pas possible de faire directement une restauration RMAN sur une base Oracle RDS.
\nIl faudra passer par le processus d\u00e9di\u00e9 qu’Amazon met en place via des backup automatiques : par exemple, l’outil AWS CLI utilisera pour cela la commande :<\/p>\n
aws rds restore-db-instance-to-point-in-time …..<\/em><\/strong><\/span><\/p>\nMais alors pourquoi effectuer une sauvegarde RMAN alors que AWS ne sait la restaurer sur une autre instance Oracle RDS ?
\nPlusieurs scenarii \u00e0 cela peuvent \u00eatre envisag\u00e9s :
\n– ma\u00eetriser sa propre strat\u00e9gie de backup\u00a0 (full, incremental, archivelog) et pouvoir externaliser les fichiers (sur S3 ou autres cibles externes)
\n– Utiliser les fichiers RMAN pour faire une restauration de base sur un serveur EC2 ou on-premise.
\n– Isoler un jeu de backup RMAN pour des besoins d’exercices comptables dans le cadre d’une compliance demand\u00e9 par un cabinet d’audit
\n…..<\/p>\n
<\/p>\n
Sauvegardes RMAN sous Oracle RDS<\/h2>\nPre requis<\/span><\/h4>\n\n- Les param\u00e8tres<\/li>\n<\/ul>\n
De nombreux param\u00e8tres utilis\u00e9s par la proc\u00e9dure rdsadmin.rdsadmin_rman_util\u00a0 <\/span><\/strong><\/span>seront communs \u00e0 chaque appel. La liste des param\u00e8tres sera la suivante :<\/p>\n\u2022 p_owner<\/strong> <\/span>: Propri\u00e9taire du r\u00e9pertoire sp\u00e9cifi\u00e9 dans p_directory_name.
\n\u2022 p_directory_name<\/strong> <\/span>: Nom de r\u00e9pertoire de la base de donn\u00e9es.
\n\u2022 p_label\u00a0<\/strong> <\/span>: nom donn\u00e9 pour le backup
\n\u2022 p_compress\u00a0<\/strong> <\/span>: Active ou non la compression (TRUE\/FALSE)
\n\u2022 p_include_archive_logs\u00a0<\/strong> <\/span>: sauvegarde avec archivelogs (TUE\/FALSE)
\n\u2022 p_include_controlfile<\/strong> <\/span>: sauvegarde avec controlfile (TRUE\/FALSE)
\n\u2022 p_parallel<\/strong> <\/span>: sauvegarde sur plusieurs canaux (nombre \u00e0 sp\u00e9cifier)
\n\u2022 p_rman_to_dbms_output<\/strong> <\/span>: affiche les instructions RMAN \u00e0 l\u2019\u00e9cran via DBMS_OUTPUT (TRUE\/FALSE). Sp\u00e9cifier \u00e9galement “set serverouput on”
\n\u2022 p_section_size_mb<\/strong> <\/span>: taille de chaque fichier de backup (correspond au MAXPIECESIZE dans RMAN)
\n\u2022 p_validation_type<\/strong> <\/span>: recherche de corruption de blocs bases de donn\u00e9es. Specifier “PHYSICAL” (par d\u00e9faut) ou “PHYSICAL+LOGICAL”<\/p>\n <\/p>\n
\n- Le DIRECTORY<\/li>\n<\/ul>\n
Avant de commencer \u00e0 effectuer une sauvegarde RMAN, il faudra s’assurer de disposer d’un DIRECTORY base de donn\u00e9es dans lequel stock\u00e9 les fichiers RMAN.
\nCelui ci pourra se cr\u00e9er via la commande :<\/p>\n
SQL> exec rdsadmin.rdsadmin_util.create_directory(p_directory_name => 'BACKUP_RMAN');<\/span>\r\n\r\nPL\/SQL procedure successfully completed.<\/span>\r\n\r\nSQL> select OWNER, DIRECTORY_NAME,DIRECTORY_PATH from dba_directories where DIRECTORY_NAME='BACKUP_RMAN';<\/span>\r\n\r\nOWNER DIRECTORY_NAME DIRECTORY_PATH<\/span>\r\n---------- ------------------------- ---------------------------------------------<\/span>\r\nSYS BACKUP_RMAN \/rdsdbdata\/userdirs\/01<\/span><\/pre>\n <\/p>\n
\n- G\u00e9rer le mode ARCHIVELOG de la base<\/li>\n<\/ul>\n
V\u00e9rifier que la base est bien en mode ARCHIVELOG et que la destination des archivelogs ne comporte pas d’erreurs :<\/p>\n
SQL>\u00a0 select NAME,CREATED,LOG_MODE,OPEN_MODE,DATABASE_ROLE from v$database;<\/span>\r\n\r\nNAME\u00a0\u00a0\u00a0\u00a0\u00a0 CREATED\u00a0\u00a0 LOG_MODE\u00a0\u00a0\u00a0\u00a0 OPEN_MODE\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 DATABASE_ROLE<\/span>\r\n--------- --------- ------------ -------------------- ----------------<\/span>\r\nCAPDATA\u00a0\u00a0 20-JUN-19 ARCHIVELOG\u00a0 \u00a0READ WRITE\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 PRIMARY<\/span>\r\n\r\n\r\nSQL> select DEST_ID,DEST_NAME,STATUS,TARGET,NAME_SPACE,DESTINATION,PROCESS,ERROR from v$archive_dest where status='VALID';<\/span>\r\n\r\n \u00a0 DEST_ID DEST_NAME\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 STATUS\u00a0\u00a0\u00a0 TARGET\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 NAME_SP DESTINATION\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 \u00a0\u00a0\u00a0 \u00a0PROCESS\u00a0\u00a0\u00a0 ERROR<\/span>\r\n---------- -------------------- --------- ---------------- ------- ---------------------------------------- ---------- ----------<\/span>\r\n\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 1 LOG_ARCHIVE_DEST_1\u00a0\u00a0 VALID\u00a0\u00a0\u00a0\u00a0 PRIMARY\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 SYSTEM\u00a0 \/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog\u00a0 \u00a0\u00a0\u00a0ARCH<\/span><\/pre>\n <\/p>\n
D\u00e9finir la r\u00e9tention des archivelogs avant de lancer une sauvegarde. Par exemple, nous pourrons conserver 12 heures d’archivelogs sur disque afin de pouvoir les exploiter par un autre processus (LogMinner par exemple).<\/p>\n
SQL> exec rdsadmin.rdsadmin_util.set_configuration( name => 'archivelog retention hours', value => '12');<\/span>\r\n\r\nPL\/SQL procedure successfully completed.<\/span>\r\n\r\nSQL> commit;<\/span>\r\n\r\nCommit complete.<\/span>\r\n\r\nSQL> exec rdsadmin.rdsadmin_util.show_configuration;<\/span>\r\nNAME:archivelog retention hours<\/span>\r\nVALUE:12<\/span>\r\nDESCRIPTION:ArchiveLog expiration specifies the duration in hours before archive\/redo log files are automatically deleted.<\/span>\r\nNAME:tracefile retention<\/span>\r\nVALUE:10080<\/span>\r\nDESCRIPTION:tracefile expiration specifies the duration in minutes before tracefiles in bdump are automatically deleted.<\/span>\r\n\r\nPL\/SQL procedure successfully completed.<\/span><\/pre>\n\n- Les versions compatibles<\/li>\n<\/ul>\n
Noter que pour effectuer des backup RMAN sur une base de donn\u00e9es Oracle RDS, vous devez disposer d’une base ayant une des versions AWS suivantes :<\/p>\n
- 11.2.0.4.v19 ou versions 11.2 sup\u00e9rieures\r\n\r\n- 12.1.0.2.v15 ou versions 12.1 sup\u00e9rieures\r\n\r\n- 12.2.0.1.ru-2019-01.rur-2019-01.r1 ou versions 12.2 sup\u00e9rieures\r\n\r\n<\/pre>\nSauvegarde compl\u00e8te<\/span><\/h4>\nPour effectuer une sauvegarde de type FULL DATABASE, nous utiliserons la proc\u00e9dure “rdsadmin.rdsadmin_rman_util.backup_database_full<\/span>“.
\nLa commande ci dessous est un exemple de backup Full de notre base CAPDATA sur le nouveau repertoire ‘BACKUP_RMAN’ cr\u00e9\u00e9 auparavant.<\/p>\nset serveroutput on\r\nBEGIN\r\n rdsadmin.rdsadmin_rman_util.backup_database_full(\r\n p_owner => 'SYS', \r\n p_directory_name => 'BACKUP_RMAN',\r\n p_parallel => 2, \r\n p_section_size_mb => 50,\r\n p_rman_to_dbms_output => TRUE);\r\nEND;\r\n\/<\/span><\/pre>\nle mode “serveroutpt on” combin\u00e9 \u00e0 l’option “p_rman_to_dbms_output” nous permet d’obtenir le mode verbeux du processus, notamment la commande exacte lanc\u00e9e par RMAN qui est :<\/p>\n
RUN_RMAN_CMD: \/rdsdbbin\/oracle\/bin\/rman TARGET \/ LOG \/rdsdbdata\/log\/trace\/rds-rman-backup-database-2019-06-21.08-59-25.417855000.txt<\/span>\r\n@\/rdsdbdata\/tmp\/rds-rman-backup-database-2019-06-21.08-59-25.417855000.input<\/span>\r\nRecovery Manager: Release 12.1.0.2.0 - Production on Fri Jun 21 08:59:29 2019<\/span>\r\nCopyright (c) 1982, 2014, Oracle and\/or its affiliates. All rights reserved.<\/span>\r\nconnected to target database: CAPDATA (DBID=703110999)<\/span>\r\nRMAN> CONFIGURE CONTROLFILE AUTOBACKUP ON;<\/span>\r\n2> CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '\/rdsdbdata\/userdirs\/01\/BACKUP-2019-06-21-08-59-25-%F';<\/span>\r\n3> CONFIGURE BACKUP OPTIMIZATION ON;<\/span>\r\n4> RUN {<\/span>\r\n5> ALLOCATE CHANNEL d1 DEVICE TYPE DISK FORMAT '\/rdsdbdata\/userdirs\/01\/BACKUP-2019-06-21-08-59-25-backup-%T-%U';<\/span>\r\n6> ALLOCATE CHANNEL d2 DEVICE TYPE DISK FORMAT '\/rdsdbdata\/userdirs\/01\/BACKUP-2019-06-21-08-59-25-backup-%T-%U';<\/span>\r\n7> crosscheck archivelog all;<\/span>\r\n8> BACKUP DATABASE SECTION SIZE 50M;<\/span>\r\n9> RELEASE CHANNEL d1;<\/span>\r\n10> RELEASE CHANNEL d2;<\/span>\r\n11> }<\/span>\r\n12><\/span><\/pre>\nNous pourrons lire dans le Directory pour voir si les fichiers sont bien pr\u00e9sents :<\/p>\n
SQL> select * from table (rdsadmin.rds_file_util.listdir(p_directory => ‘BACKUP_RMAN’));<\/span><\/p>\nFILENAME TYPE FILESIZE MTIME<\/span>\r\n------------------------------------------------------------ ---------- ---------- -------------------<\/span>\r\n01\/ directory 4096 2019\/06\/21 09:09:33<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_4_1 file 49152 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_4_1 file 45965312 2019\/06\/21 08:59:35<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_6_1 file 57344 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_1_1 file 35233792 2019\/06\/21 08:59:33<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_6_1 file 45842432 2019\/06\/21 08:59:41<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_6_1 file 49684480 2019\/06\/21 08:59:36<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0mu4l1nu_2_1 file 49152 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_5_1 file 90112 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_7_1 file 38756352 2019\/06\/21 08:59:37<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0ou4l1nu_1_1 file 6463488 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_1_1 file 2351104 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_3_1 file 42885120 2019\/06\/21 08:59:39<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_8_1 file 22667264 2019\/06\/21 08:59:38<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_3_1 file 49152 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_2_1 file 35938304 2019\/06\/21 08:59:38<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_2_1 file 33734656 2019\/06\/21 08:59:34<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_7_1 file 19382272 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_5_1 file 48603136 2019\/06\/21 08:59:36<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0mu4l1nu_1_1 file 6356992 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_4_1 file 37306368 2019\/06\/21 08:59:40<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-0gu4l1nu_2_1 file 573440 2019\/06\/21 08:59:42<\/span>\r\nBACKUP-2019-06-21-08-59-25-c-703110999-20190621-00 file 8421376 2019\/06\/21 08:59:43<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-02u4l1nk_5_1 file 52469760 2019\/06\/21 08:59:40<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_3_1 file 28262400 2019\/06\/21 08:59:34<\/span>\r\nBACKUP-2019-06-21-08-59-25-backup-20190621-01u4l1nk_1_1 file 25346048 2019\/06\/21 08:59:33<\/span><\/pre>\nOn pourra \u00e9galement interroger la vue “v$rman_backup_job_details” habituelle pour v\u00e9rifier la prise en compte de notre sauvegarde dans le controlfile :<\/p>\n
SQL> select START_TIME,END_TIME,INPUT_BYTES\/1024\/1024,OUTPUT_BYTES\/1024\/1024,STATUS,INPUT_TYPE,ELAPSED_SECONDS\/60 “Minutes”
\n<\/span>from v$rman_backup_job_details <\/span>where START_TIME > sysdate – 1 order by START_TIME;<\/span><\/p>\nSTART_TIME END_TIME INPUT_BYTES\/1024\/1024 OUTPUT_BYTES\/1024\/1024 STATUS INPUT_TYPE Minutes<\/span>\r\n------------------- ------------------- --------------------- ---------------------- ----------------------- ------------- ----------<\/span>\r\n2019\/06\/21 08:59:32 2019\/06\/21 08:59:44 741.0625 559.164063 COMPLETED DB FULL .2<\/span><\/pre>\n <\/p>\n
Sauvegarde incr\u00e9mentale<\/span><\/h4>\nIl sera possible d’effectuer un backup de type INCREMENTAL. Pour cela nous utiliserons la proc\u00e9dure “rdsadmin.rdsadmin_rman_util.backup_database_incremental<\/span>“.<\/p>\n\n- Le block change tracking<\/li>\n<\/ul>\n
Le block change tracking pourra \u00eatre activer afin d’optimiser les backup incr\u00e9mentaux, nous utiliserons la proc\u00e9dure “rdsadmin.rdsadmin_rman_util.enable_block_change_tracking”<\/span><\/p>\nSQL> exec rdsadmin.rdsadmin_rman_util.enable_block_change_tracking;<\/span>\r\n\r\nPL\/SQL procedure successfully completed.<\/span>\r\n\r\nSQL> select * from v$block_change_tracking;<\/span>\r\n\r\nSTATUS FILENAME BYTES CON_ID<\/span>\r\n---------- ------------------------------------------------------------ ---------- ----------<\/span>\r\nENABLED \/rdsdbdata\/db\/CAPDATA_A\/changetracking\/o1_mf_gjs99w4p_.chg 11599872 0<\/span><\/pre>\npar la suite, un backup incr\u00e9mental pourra \u00eatre lanc\u00e9.
\nOn oubliera pas de d\u00e9finir le param\u00e8tre “p_level” , d’abord avec un “p_level” \u00e0 0, puis un “p_level” \u00e0 1 :<\/p>\n
BEGIN<\/span>\r\n rdsadmin.rdsadmin_rman_util.backup_database_incremental(<\/span>\r\n p_owner => 'SYS', <\/span>\r\n p_directory_name => 'BACKUP_RMAN',<\/span>\r\n\tp_level\t\t => 1,<\/span>\r\n p_parallel => 2, <\/span>\r\n p_section_size_mb => 50,<\/span>\r\n p_rman_to_dbms_output => TRUE);<\/span>\r\nEND;<\/span>\r\n\/\r\n<\/span><\/pre>\nComme pour la sauvegarde FULL, il sera possible d’aller voir dans le directory BACKUP_RMAN les fichiers g\u00e9n\u00e9r\u00e9s, ainsi qu’interroger la vue v$rman_backup_job_details pour voir si notre sauvegarde incr\u00e9mentale est bien pass\u00e9e. A noter que le param\u00e8tre “p_rman_to_dbms_output” \u00e0 TRUE permet de voir directement \u00e0 l’\u00e9cran la sortie de la commande RMAN.
\nGr\u00e2ce au block change tracking activ\u00e9, on voit que, lors du passage du “p_level” 1, RMAN ne traite plus que 300Mo de blocks de donn\u00e9es en entr\u00e9e au lieu des 740Mo lors du backup full.<\/p>\n
SQL> select START_TIME,END_TIME,INPUT_BYTES\/1024\/1024,OUTPUT_BYTES\/1024\/1024,STATUS,INPUT_TYPE,ELAPSED_SECONDS\/60 \"Minutes\" from v$rman_backup_job_details<\/span>\r\n 2 where START_TIME > sysdate - 1 and INPUT_TYPE like '%INCR%' order by START_TIME;<\/span>\r\n\r\nSTART_TIME END_TIME INPUT_BYTES\/1024\/1024 OUTPUT_BYTES\/1024\/1024 STATUS INPUT_TYPE Minutes<\/span>\r\n------------------- ------------------- --------------------- ---------------------- ----------------------- ------------- ----------<\/span>\r\n2019\/06\/21 09:49:41 2019\/06\/21 09:50:06 744.40625 561.40625 COMPLETED DB INCR .416666667<\/span>\r\n2019\/06\/21 09:50:32 2019\/06\/21 09:50:55 300.820313 9.984375 COMPLETED DB INCR .383333333<\/span><\/pre>\n <\/p>\n
Sauvegarde archivelogs\u00a0<\/span><\/h4>\nPour la sauvegarde des archivelogs, nous pourrons utiliser 4 m\u00e9thodes distinctes \u00e0 savoir<\/p>\n
\n- sauvegardes des archivelogs complet<\/li>\n
- sauvegardes des archivelogs \u00e0 partir d’une plage de dates<\/li>\n
- sauvegardes des archivelogs \u00e0 partir d’une plage de SCN<\/li>\n
- sauvegardes des archivelogs \u00e0 partir d’une plage de s\u00e9quences<\/li>\n<\/ul>\n
La sauvegarde archivelogs compl\u00e8te<\/h5>\n
On lancera la proc\u00e9dure “backup_archivelog_all<\/span>” pour cette op\u00e9ration :<\/p>\nBEGIN<\/span>\r\n rdsadmin.rdsadmin_rman_util.backup_archivelog_all(<\/span>\r\n p_owner => 'SYS', <\/span>\r\n p_directory_name => 'BACKUP_RMAN',<\/span>\r\n p_parallel => 2, <\/span>\r\n p_rman_to_dbms_output => TRUE);<\/span>\r\nEND;<\/span>\r\n\/<\/span><\/pre>\nPour v\u00e9rifier, soit on regarde la liste des fichiers dans ‘BACKUP_RMAN’, soit on interroge la vue v$rman_backup_job_details.<\/p>\n
SQL> select START_TIME,END_TIME,INPUT_BYTES\/1024\/1024,OUTPUT_BYTES\/1024\/1024,STATUS,INPUT_TYPE,ELAPSED_SECONDS\/60 \"Minutes\" from v$rman_backup_job_details<\/span>\r\n 2 where START_TIME > sysdate - 1 and INPUT_TYPE like '%ARC%' order by START_TIME;<\/span>\r\n\r\nSTART_TIME END_TIME INPUT_BYTES\/1024\/1024 OUTPUT_BYTES\/1024\/1024 STATUS INPUT_TYPE Minutes<\/span>\r\n------------------- ------------------- --------------------- ---------------------- ----------------------- ------------- ----------<\/span>\r\n2019\/06\/21 10:07:55 2019\/06\/21 10:07:57 6.93212891 6.93701172 COMPLETED ARCHIVELOG .033333333<\/span><\/pre>\nla sauvegarde archivelogs avec plage de dates<\/h5>\n
On utilisera pour cela la proc\u00e9dure “backup_archivelog_date<\/span>” avec les param\u00e8tres ‘p_from_date’ et ‘p_to_date’ :<\/p>\nBEGIN<\/span>\r\n rdsadmin.rdsadmin_rman_util.backup_archivelog_date(<\/span>\r\n p_owner => 'SYS', <\/span>\r\n p_directory_name => 'BACKUP_RMAN',<\/span>\r\n p_from_date => '21\/06\/2019 09:00:00',<\/span>\r\n p_to_date => '21\/06\/2019 10:00:00',<\/span>\r\n p_parallel => 2, <\/span>\r\n p_rman_to_dbms_output => FALSE);<\/span>\r\nEND;<\/span>\r\n\/<\/span><\/pre>\nOn v\u00e9rifie :<\/p>\n
SQL> select START_TIME,END_TIME,INPUT_BYTES\/1024\/1024,OUTPUT_BYTES\/1024\/1024,STATUS,INPUT_TYPE,ELAPSED_SECONDS\/60 \"Minutes\" from v$rman_backup_job_details<\/span>\r\n 2 where START_TIME > sysdate - 1 and INPUT_TYPE like '%ARC%' order by START_TIME;<\/span>\r\n\r\nSTART_TIME END_TIME INPUT_BYTES\/1024\/1024 OUTPUT_BYTES\/1024\/1024 STATUS INPUT_TYPE Minutes<\/span>\r\n------------------- ------------------- --------------------- ---------------------- ----------------------- ------------- ----------<\/span>\r\n21\/06\/2019 10:07:55 21\/06\/2019 10:07:57 6.93212891 6.93701172 COMPLETED ARCHIVELOG .033333333<\/span>\r\n21\/06\/2019 10:20:03 21\/06\/2019 10:20:05 6.77636719 6.78076172 COMPLETED ARCHIVELOG .033333333<\/span><\/pre>\nla sauvegarde archivelogs avec plage SCN<\/h5>\n
Nous utiliserons la proc\u00e9dure “backup_archivelog_scn<\/span>” avec les param\u00e8tres principaux ‘p_from_scn’ et ‘p_to_scn’ .
\nOn recherchera, par exemple les derniers archivelogs g\u00e9n\u00e9r\u00e9s avec les SCN associ\u00e9s (avec 1 h d’historique) :<\/p>\nSQL> select NAME,FIRST_TIME,FIRST_CHANGE# from v$archived_log where FIRST_TIME > sysdate - (1\/24) order by 3;<\/span>\r\n\r\nNAME FIRST_TIME FIRST_CHANGE#<\/span>\r\n------------------------------------------------------------ ------------------- -------------<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-258-1-1011452844.arc 21\/06\/2019 11:14:42 612259<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-259-1-1011452844.arc 21\/06\/2019 11:19:42 612396<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-260-1-1011452844.arc 21\/06\/2019 11:24:42 612566<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-261-1-1011452844.arc 21\/06\/2019 11:29:42 612706<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-262-1-1011452844.arc 21\/06\/2019 11:34:40 612840<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-263-1-1011452844.arc 21\/06\/2019 11:39:40 612993<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-264-1-1011452844.arc 21\/06\/2019 11:44:40 613124<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-265-1-1011452844.arc 21\/06\/2019 11:49:40 613257<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-266-1-1011452844.arc 21\/06\/2019 11:54:40 613420<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-267-1-1011452844.arc 21\/06\/2019 11:59:41 613550<\/span>\r\n\/rdsdbdata\/db\/CAPDATA_A\/arch\/redolog-268-1-1011452844.arc 21\/06\/2019 12:04:41 613838\r\n\r\n<\/span><\/pre>\nA partir de la, on sera libre de prendre une plage de SCN :<\/p>\n
BEGIN<\/span>\r\n rdsadmin.rdsadmin_rman_util.backup_archivelog_scn(<\/span>\r\n p_owner => 'SYS', <\/span>\r\n p_directory_name => 'BACKUP_RMAN',<\/span>\r\n p_from_scn => 612396,<\/span>\r\n p_to_scn => 613550,<\/span>\r\n p_parallel => 2, <\/span>\r\n p_rman_to_dbms_output => TRUE);<\/span>\r\nEND;<\/span>\r\n\/<\/span><\/pre>\nOn v\u00e9rifie<\/p>\n
SQL> select START_TIME,END_TIME,INPUT_BYTES\/1024\/1024,OUTPUT_BYTES\/1024\/1024,STATUS,INPUT_TYPE,ELAPSED_SECONDS\/60 \"Minutes\" from v$rman_backup_job_details<\/span>\r\n 2 where START_TIME > sysdate - 1 and INPUT_TYPE like '%ARC%' order by START_TIME;<\/span>\r\n\r\nSTART_TIME END_TIME INPUT_BYTES\/1024\/1024 OUTPUT_BYTES\/1024\/1024 STATUS INPUT_TYPE Minutes<\/span>\r\n------------------- ------------------- --------------------- ---------------------- ----------------------- ------------- ----------<\/span>\r\n21\/06\/2019 10:07:55 21\/06\/2019 10:07:57 6.93212891 6.93701172 COMPLETED ARCHIVELOG .033333333<\/span>\r\n21\/06\/2019 10:20:03 21\/06\/2019 10:20:05 6.77636719 6.78076172 COMPLETED ARCHIVELOG .033333333<\/span>\r\n21\/06\/2019 12:16:56 21\/06\/2019 12:16:58 2.36523438 2.36816406 COMPLETED ARCHIVELOG .033333333<\/span><\/pre>\n <\/p>\n
la sauvegarde archivelogs avec plage de sequences<\/h5>\n
Comme pour la date ou le SCN, le backup RMAN avec plage de sequences s’appuiera sur un num\u00e9ro de s\u00e9quence d’archivelog g\u00e9n\u00e9r\u00e9. La proc\u00e9dure “backup_archivelog_sequence<\/span>” utilisera les param\u00e8tres ‘p_from_sequence’ et ‘p_to_sequence’.<\/p>\n
![\"twitter\"](\"https:\/\/blog.capdata.fr\/wp-content\/plugins\/social-media-feather\/synved-social\/image\/social\/regular\/48x48\/twitter.png\" "\\"Share")
<\/a>![\"linkedin\"](\"https:\/\/blog.capdata.fr\/wp-content\/plugins\/social-media-feather\/synved-social\/image\/social\/regular\/48x48\/linkedin.png\" "\\"Share")
<\/a>![\"mail\"](\"https:\/\/blog.capdata.fr\/wp-content\/plugins\/social-media-feather\/synved-social\/image\/social\/regular\/48x48\/mail.png\" "\\"Share")
<\/a>![\"\"](\"https:\/\/blog.capdata.fr\/wp-content\/uploads\/2018\/12\/aws-database.png\")
AWS, dans son offre RDS, emp\u00eache l’utilisateur d\u2019acc\u00e9der au serveur ou est h\u00e9berg\u00e9e la base de donn\u00e9es. Ce qui fait que l’utilisation de binaires type “expdp”, “dgmgrl” ou encore “rman” ne sera pas possible sur le serveur. Par contre, AWS autorisera l’utilisateur \u00e0 effectuer des backup RMAN directement en base.<\/p>\n