We have a oracle 10.2 db running on a two node cluster (RAC). we use data protector 6.0 to backup the db.frequently the data protector throws errors saying[code][Major] From: OB2BAR_DMA@db-odbs1.db.ae "dbgold" Time: 9/1/2010 11:16:11 PMCannot get information on file '/var/opt/omni/tmp/ctrl_dbgold.dbf' Error: 2.[Normal] From: OB2BAR_DMA@db-odbs1.db.ae "dbgold" Time: 9/1/2010 11:16:48 PMCompleted OB2BAR Backup: db-odbs1:dbgold DP Managed Control File Backup "Oracle8"[Major] From: firstname.lastname@example.org "dbgold" Time: 09/01/10 23:16:48Backup of Data Protector managed control file backup failed.[/code]Can someone throw more light on this please?
Well - I ran in the same problem and it is definitly an oracle problem.
The point is if you commit a
alter DATABASE BACKUP CONTROLFILE TO '/usr/omni/tmp/ctrl_.dbf' REUSE;"
and this is what HPDP scripts do here,
in an RAC envirnment, you do not know on which RAC-node this command will be executed.
Actually with Oracle 11g R2 ( not R1) the Oracle Documentation changed and said, that if you want to do a DATABASE BACKUP CONTROLFILE on a RAC, you need to use a shared storage.
And that's the way to solve the problem !
We have disabled (clicked on DP checkbox) DP controlled control file backup and use "the RMAN" one.
Works fine, even recovery from bare metal brought the system up in the most current state.