SVR4 Error?

RMAN-06403: could not obtain a fully authorized session
ORA-01034: ORACLE not available
ORA-27101: shared memory realm does not exist
SVR4 Error: 2: No such file or directory

when the backup trying to backup it gives this error in log?

why?

Thanks
Pat

a) You did not set the ORACLE_HOME/ORACLE_SID

b) The database is not started
:rolleyes:

oracle_home/oracle_sid is set in the script of backup and when i do ps -ef|grep datatabase it does return records so that means database is running? please let me know if there is other way to check it.

OS: sun solaris:
Oracle 10.1.0.3.0

Restart DB if required. this may solve the problem...
cheers/
MR

This may provide a clue:

SVR4 Error: 2: No such file or directory

Show us your script. :cool:

#!/bin/ksh
######################################################################
# Standard RMAN Script to backup the controlfile to Disk.
#
# ?? ??? 03 ????? Script Created
# ?? ??? 03 Sunny Script Modified
# a. Prevent multiple instances of the same
# script from running.
# b. Use Parameters file for rman catalog,
# password, log file names, etc.
# ?? Mar 06 Sunny Script Modified
# a. Common routines written to a separate script.
######################################################################
umask 002
typeset -i STATUS
typeset -i RETCODE
export STATUS=0
export RETCODE=0
BACKUP_TYPE=cntl
######################################################################
# Source the file containing the common functions
######################################################################
. /app/oracle/admin/scripts/rman/RMAN_common_functions.sh
LOGNAME=$CNTLLOG
## Determine ORACLE HOME, LD_LIBRARY_PATH.
get_oracle_home
echo `date` "Begin $BACKUP_TYPE backup of $ORACLE_SID" >> $SUMMARYLOG
echo `date` "Begin $BACKUP_TYPE backup of $ORACLE_SID" >> $SERVERLOG

#######################################################################
# Ensure that another instance of this backup is not already running
#######################################################################
cd /app/oracle/admin/scripts/rman
# PROGRAM = RMAN script name minus "RMAN_" and the ".sh"
PROGRAM=`echo $0|sed s/'RMAN_'/''/g|sed s/'.sh'/''/g|sed s/'\/app\/oracle\/admin\/scripts\/rman\/'/''/g|sed s/'.\/'/''/g`
CUR_DATE=`date "+%b%d.%H:%M:%S"`
OEM_LOG=$HOME/logs/oem_errors.log
check_duplicate_process
touch $LOCKFILE

######################################################################
# Delete old backup files based on retention times specified
# (as parameters)
######################################################################
cleanup_backup_files

######################################################################
# Start the rman backup.
######################################################################
echo `date` "Start rman script for ($BACKUP_TYPE) backup of $ORACLE_SID" >> $SUMMARYLOG
echo `date` "Start rman script for ($BACKUP_TYPE) backup of $ORACLE_SID" >> $SERVERLOG
$ORACLE_HOME/bin/rman target / >> $LOGNAME < connect $CATCMD $RMAN_USER/$RMAN_PASSWORD@$RCAT_SID;
run {execute script backup_cf_to_disk;}
EOF
echo `date` "End rman script for ($BACKUP_TYPE) backup of $ORACLE_SID" >> $SUMMARYLOG
echo `date` "End rman script for ($BACKUP_TYPE) backup of $ORACLE_SID" >> $SERVERLOG

######################################################################
# Check for errors.
######################################################################
if [ `grep "RMAN-00569" $LOGNAME |wc -l` -ne 0 ]
then
(( STATUS = STATUS | 4 ))
send_error_message
fi
######################################################################
# Identify files to be saved.
# From the logfile, identify all the backup pieces created during
# this backup. These files will be compressed (if required) and saved.
######################################################################

if [ $SAVE_FILES_YN = "Y" ]
then
if [ $COMPRESS_YN = "Y" ]
then
cat $LOGNAME|grep "piece handle" |$AWK -F= '{print $2}'|$AWK '{print $1".gz"}'> /app/oracle/admin/scripts/rman/${PROGRAM}.${CUR_DATE}.${ORACLE_SID}.sa
ve_list
else
cat $LOGNAME|grep "piece handle" |$AWK -F= '{print $2}'|$AWK '{print $1}'> /app/oracle/admin/scripts/rman/${PROGRAM}.${CUR_DATE}.${ORACLE_SID}.save_li
st
fi
fi

######################################################################
# Compress / gzip the backup files
######################################################################
if [ $COMPRESS_YN = "Y" ]
then
gzip_files
fi

######################################################################
# Save backup files (tape).
# Delete the lockfile. This allows additional backups of the same time
# to run even before the current save completes.
######################################################################
rm $LOCKFILE
if [ $SAVE_FILES_YN = "Y" ]
then
save_files &
fi

######################################################################
# Cleanup and exit
######################################################################
echo `date` "End $BACKUP_TYPE backup of ${ORACLE_SID}" >> $SUMMARYLOG
echo `date` "End $BACKUP_TYPE backup of ${ORACLE_SID}" >> $SERVERLOG
exit 0

hot backup is fine but not think controlfile backup is having problem!! please guide!!

Is this the first time you try and execute this script?

--- DELETED ---

Ooops, the typo I found is actually generated by the forum...:p

no it's been running since long and just started show error in email like follow:
RMAN-06403: could not obtain a fully authorized session
ORA-01034: ORACLE not available
ORA-27101: shared memory realm does not exist
SVR4 Error: 2: No such file or directory

Long shot: check if your Oracle home mount point is 100% full.
:confused:

Hi how can i check that please guide i am totally confuse!!

On solaris command line do:

Code:

---

$ df -h

---

and see if there is an entry of 100% under the use% label.
;)
PS: if you see 100% for the Oracle home path, you need to make space by removing obsolete files/logs/etc...

it doesn't work:

$df -h
df: unknown option: h
Usage: df [-F FSType] [-abegklntVv] [-o FSType-specific_options] [directory | block_device | resource]

come on - basic unix!

after running df -k i get this:

oracle-eis-u01
512000 150325 339076 31% /oracle/eis/u01

so i think it is not 100% at all!!

Quote:

---

Originally Posted by midora

after running df -k i get this:
oracle-eis-u01
512000 150325 339076 31% /oracle/eis/u01
so i think it is not 100% at all!!

---

Is there where your Oracle home is?
Any other mount point at 100%?

PS: Must be old version of Solaris to not recognize the 'h' option.

Okay... lets go back to the basics and troubleshoot this properly.

I assume you are connecting to a local database, am I right?

Log into the box as "oracle" then...

1- Set your instance e.g.: OH and SID

2- Do and post...
ps -ef | grep pmon

3- Do and post...
sqlplus /nolog
conn sys as sysdba
select instance_name, status from v$instance;

4- Post or attach your whole RMAN log, not only the three lines you have posted.

sure wll do that shortly!!

thanks
Pat