In Oracle Enterprise Manager( OEM ) 12c and 13c, there are a couple of newly added cluster databases targets and database system targets are in “PENDING” status, though all the cluster database instances are showing up.
Read more of this content when you subscribe today.
“_rm_dup_.dat” file is for “restore optimization” purpose to record which datafiles have been restored/copied onto auxiliary host, when running RMAN “duplicate target database for standby from active database”.
There is a file with name “_rm_dup_$ORACLE_SID.dat” under $ORACLE_HOME/dbs of the clone database. This file is from executing “duplicate database ….”
“_rm_dup_<dup_db>.dat” stores the names of datafilecopy already created by duplicate. Inside this file, rman can find the name of the datafiles already copied to auxiliary host. if, for some reasons, the duplicate failed, the second duplicate will check this file. If the datafilecopy has been created on auxilary host, and the vital information of the datafile (file number, database id, creation scn, database name) and its checkpoint scn is behind the until scn, then the datafilecopy can be used by this new duplicate and restore/copy is not necessary.
So this file helps for “restore optimization” purpose when running RMAN “duplicate database …..”.
It is a good practice to temporarily disable CRS auto start up when OS upgrading or patching.
Read more of this content when you subscribe today.
In GI/RAC environment, directories and files permission should be identical on all nodes.
When non-oracle user ran “srvctl status service -d RACTEST” on second node, get the following errors. But it did not happen on the first node.
$ srvctl status service -d RACTEST sclsdglnam failed path: /u01/app/oracle/product/11.2.0/dbhome_2/log/racnode2/client filename: clsc 2016-08-31 11:57:42.820: CLSD: An error occurred while attempting to generate a full name. Logging may not be active for this process Additional diagnostics: CLSU-00100: Operating System function: access failed failed with error data: 13 CLSU-00101: Operating System error message: Permission denied CLSU-00103: error location: SlfAccess (:CLSD00183:) category: -3 operation name: access failed location: SlfAccess dependent info: 13
There is a file/directory permission issue on the second node.
$cd $ORACLE_HOME/log $ls -ltr .. .. drwxr----T 3 oracle oinstall 4096 Aug 16 16:34 racnode2 $cd racnode2 $ls -l drwxr----T 2 oracle oinstall 4096 Aug 31 09:58 client
While on node 1, the file permission is different.
$cd ORACLE_HOME/log $ls -ltr drwxr-xr-t 3 oracle oinstall 4096 Aug 16 14:13 racnode1 drwxr-xr-x 3 oracle oinstall 4096 Aug 20 17:10 diag $cd racnode1 $ls -ltr total 4 drwxr-xr-t 2 oracle oinstall 4096 Aug 26 12:46 client
The solution is to change the directory/file permissions on node 2 to make it same as on node 1.
$cd $ORACLE_HOME/log $chmod 1755 racnode2 $cd racnode2 $ls -ltr total 4 drwxr----T 2 oracle oinstall 4096 Aug 31 09:58 client $chmod 1755 client $ ls -ltr total 4 drwxr-xr-t 2 oracle oinstall 4096 Aug 31 09:58 client
It is recommended to relink GI and RAC Home binaries after OS upgrading or patching.
This post demonstrates how to relink Oracle Grid Infrastructure ( GI ) and RAC Home binaries after OS upgrading or patching.
Read more of this content when you subscribe today.
Make DBA Life Easy 