| validate RMAN backup [message #324079] |
Fri, 30 May 2008 12:07  |
simoh
Messages: 6
Registered: May 2008
|
Junior Member |
|
|
Hi,
I really stuck and need help. I have been working on this for the last couple of weeks.
This is 9.2.0.5.0 RAC environment. RMAN is used to take the backup. The database is running fine and backup is taken using cron 3 times a week. There is no error messages generated when the backup is taken.
The backup script is divided into 2 scripts. One for a full backup and the other for archive log files:
==================================================
resync catalog;
report schema;
backup database;
delete noprompt obsolete;
crosscheck archivelog all;
delete noprompt expired archivelog all;
crosscheck backupset;
delete noprompt expired backupset;
list backupset summary;
==================================================
resync catalog;
backup archivelog all delete input;
================================
However, in the alert log file I noticed that there are corrupted blocks in the backup files. I tried to validate the backup using (restore database validate) but I ended up with erros. Then, I tried (validate backupset) to check the corruption and I am getting different errors. 2 of the backupsets says that the data file can not be verfied. 1 of the backupset throws ORA-06510: PL/SQL: unhandled user-defined exception
here is a snapshot of the file:
Tue May 27 09:05:14 2008
corrupt block 3424264 found during reading backup piece,file=/racdb1/orabackup/RACDB1_kcjhctb3_1_1.bkup, corr_type=-2
Reread of blocknum=3424264, file=/racdb1/orabackup/RACDB1_kcjhctb3_1_1.bkup, found same data
there are corrputed blocked up to 3424383.
I ran another script to combine the 2
======================================
resync catalog;
report schema;
backup database plus archivelog;
delete noprompt obsolete;
crosscheck archivelog all;
delete noprompt expired archivelog all;
crosscheck backupset;
delete noprompt expired backupset;
list backupset summary;
restore database validate;
======================================
and got the same error:
corrupt block 2678111 found during reading backup piece,file=/racdb1/orabackup/RACDB1_lbjhi623_1_1.bkup, corr_type=-2
Reread of blocknum=2678111, file=/racdb1/orabackup/RACDB1_lbjhi623_1_1.bkup, found same data
I checked individual backupsets and got the following errors for 2 backupsets only:
1)ora-27069: skgfdisp: attempt to do I/O beyond the range of the file
2)ora-06510: PL/SQL: unhandled user-defined exception
does that mean the script is running fine and the problem is from the file system?
Please note that the filesystem that holds the backup uses SAMFS (veritas) to take the cashed data from the filesystem and transfer it to tapes. Also, i checked the v$database_block_corruption view and it doesnt have any data.
Thank you in advance
|
|
|
|
|
|
| Re: validate RMAN backup [message #324082 is a reply to message #324081] |
Fri, 30 May 2008 12:42  |
simoh
Messages: 6
Registered: May 2008
|
Junior Member |
|
|
Thank you for the kind reply
Are you using the autoextend on in database?
yes, some of the data files are autoextend
What is the size of the said backup piece?
What is your RMAN configuration? Any limits on the filesize?
RMAN configuration parameters are:
CONFIGURE RETENTION POLICY TO REDUNDANCY 1; # default
CONFIGURE BACKUP OPTIMIZATION ON;
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # default
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '/racdb1/orabackup/%F';
CONFIGURE DEVICE TYPE DISK PARALLELISM 4;
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT '/racdb1/orabackup/%d_%U.bkup';
CONFIGURE CHANNEL 2 DEVICE TYPE DISK FORMAT '/racdb1/orabackup/%d_%U.bkup';
CONFIGURE CHANNEL 3 DEVICE TYPE DISK FORMAT '/racdb1/orabackup/%d_%U.bkup';
CONFIGURE CHANNEL 4 DEVICE TYPE DISK FORMAT '/racdb1/orabackup/%d_%U.bkup';
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/opt/oracle/product/9.2.0/dbs/snapcf_racdb1n2.f'; # default
Also certain filesystems may need the "largefile Option" to be turned on. Check with your sysadmin/storage admin.
i really cant answer this question. as far as i understood from the OS team, it is not really a file system. the data is cashed-using RMAN- to the logical file system /racdb1/orabackup then SAMFS will send the data to tapes. is there any command that i can use to verify that?
appreciate the help
|
|
|
|
|
|
] 
Blog
Search
Help
Members
Register
Login
Home
