ORACLE_HOME is shared by both RAC instance A and RAC instance B. Both instance A and B are shut down, then ORACLE_HOME is patched, the ORACLE_HOME/bin/oracle binary will be as below just after the patching:
$ id oracle
uid=122(oracle) gid=202(oinstall) groups=202(oinstall),201(dba)
$ id grid
uid=518(grid) gid=202(oinstall) groups=202(oinstall),201(dba)
$ ls -ltr /dev/oracleasm/disks/
total 0
brw-rw---- 1 grid dba 253, 16 Jul 30 16:38 OCR_VOTE01
brw-rw---- 1 grid dba 253, 50 Jul 30 16:38 ASM_FRA01
brw-rw---- 1 grid dba 253, 57 Jul 30 16:38 ASM_DISK01
..
.
$ ls -ltr /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
-rwsr-s--x 1 oracle oinstall 228655023 May 17 11:11 /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
Manually start up instance A by using sqlplus without using ‘srvctl start instance’, and check oracle binary gid( oinstall) still the same:
$ sqlplus / as sysdba ... .. . Connected to an idle instance. SQL> startup ORACLE instance started. ... .. . Database mounted. Database opened. SQL> exit $
$ ls -ltr /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
-rwsr-s--x 1 oracle oinstall 228655023 May 17 11:11 /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
Then start up instance B through clusterware by using ‘srvctl start ‘, and noticed clusterware changed the group of the oracle binary:
$srvctl start database -d TESTB
$ ls -ltr /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
-rwsr-s--x 1 oracle dba 228655023 May 17 11:11 /u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle
The instance A got below errors in trace files and alert log:
Alert.log:
Process m000 died, see its trace file Wed Jul 25 16:37:42 2018 Process m000 died, see its trace file .. . Process q000 died, see its trace file Wed Jul 25 16:41:54 2018 Process W000 died, see its trace file Process W000 died, see its trace file
Trace file:
$ cat TESTA1_m000_7813.trc .. . /u01/app/oracle/diag/rdbms/testa/TESTA1/trace/TESTA1_m000_7813.trc ORACLE_HOME = /u01/app/oracle/product/11.2.0/dbhome_1 *** 2018-07-25 16:35:55.662 Died during process startup with error 27140 (seq=39770) OPIRIP: Uncaught error 27140. Error stack: ORA-27140: attach to post/wait facility failed ORA-27300: OS system dependent operation:invalid_egid failed with status: 1 ORA-27301: OS failure message: Operation not permitted ORA-27302: failure occurred at: skgpwinit6 ORA-27303: additional information: startup egid = 202 (oinstall), current egid = 201 (dba)
crsd_oraagent_oracle trace file shows clusterware is changed RAC $ORACLE_HOME/bin/oracle binary by using “setasmgidwrap” while instance B is started up by using ‘srvctl start ‘ by clusterware.
crsd_oraagent_oracle_42.trc:2018-07-25 16:34:31.584046 :CLSDYNAM:2785666816: [ora.testb.db]{1:44997:10712} [start] command = '/u01/app/12.1.0.2/grid/bin/setasmgidwrap oracle_binary_path=/u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle'
crsd_oraagent_oracle_42.trc:2018-07-25 16:34:31.584068 :CLSDYNAM:2785666816: [ora.testb.db]{1:44997:10712} [start] Utils:execCmd action = 1 flags = 6 ohome = /u01/app/12.1.0.2/grid cmdname = setasmgidwrap.
crsd_oraagent_oracle_42.trc:2018-07-25 16:43:38.726230 :CLSDYNAM:2128570112: [ora.testb.db]{1:44997:10835} [start] command = '/u01/app/12.1.0.2/grid/bin/setasmgidwrap oracle_binary_path=/u01/app/oracle/product/11.2.0/dbhome_1/bin/oracle'
crsd_oraagent_oracle_42.trc:2018-07-25 16:43:38.726270 :CLSDYNAM:2128570112: [ora.testb.db]{1:44997:10835} [start] Utils:execCmd action = 1 flags = 6 ohome = /u01/app/12.1.0.2/grid cmdname = setasmgidwrap.
CONCLUSION:
Always shutdown or startup RAC instances by using ‘srvctl start” through clusterware. Otherwise you might get ORA-27140 ORA-27300 ORA-27301 ORA-27302 ORA-273003 errors.
Make DBA Life Easy 