Tuesday, March 1, 2016

ORA-00845: MEMORY_TARGET not supported on this system

MEMORY_TARGET not supported on this system

Applies to:
            Oracle Enterprise Linux – Version: 5
            Oracle Database 11gR2

Description:
            I am going to create a new TEST instance from Production database. Hardware resources on TEST server less than from Production server. After restore the backup when I am going to start the database I got the following error message:

SQL> connect / as sysdba
Connected to an idle instance.
SQL> startup;
ORA-00845: MEMORY_TARGET not supported on this system

Solution:
After some research work I have found the following solution:
·         Check out /dev/shm is properly mounted by using following commands:
#df -h or #df -k command
·         Size of /dev/shm should be greater than MEMORY_TARGET or MEMMORY_MAX_TARGET parameter value
·         We see that /dev/shm have assigned 2G memory, and MEMORY_TARGET more than 2G then above ORA-845 will arise.
·         If you have MEMORY_TARGET or MEMORY_MAX_TARGET set to 4G then you should mount shared memory to greater than 4G.
·         Now login to root user and execute the following commands to fix the error:
# umount tmpfs
# mount -t tmpfs shmfs -o size=5G /dev/shm
·         In order to make the settings persistence you should add an entry in /etc/fstab similar to the following:
shmfs /dev/shm tmpfs size=5G 0


Your comments, especially which will help us improve the functionality, will be greatly appreciatedJ


Tuesday, June 16, 2015

Login Page Blank EBS R12

Blank EBS R12 Login page

Applies to:
            Oracle Enterprise Linux – Version: 5
            Oracle EBS R12

Description:
            After cloning EBS R12 successfully, when I am going to login EBS clone it did not show anything on the page. Login page is totally blank. I have restarted all the services of application and database but the problem is same. I have also restarted the whole machine/server but still the same issue.

Solution:
Following the steps to resolve the above issue:
1-      Stop all application services          
$INST_TOP/admin/scripts/adstpall.sh

2-      Remove any lock files from following folders:
$INST_TOP/ora/10.1.3/j2ee/oacore/persistence/oacore_default_group_1
$INST_TOP/ora/10.1.3/j2ee/oafm/persistence/oafm_default_group_1
$INST_TOP/ora/10.1.3/j2ee/forms/persistence/forms_default_group_1

3-      Compile the JSP files with the following command
cd $FND_TOP/patch/115/bin
perl ojspCompile.pl --compile --flush -p 2

4-      Run Autoconfig first on the DB Tier, and then on APPS Tier

5-      Start appl services
$INST_TOP/admin/scripts/adstrtall.sh



Your comments, especially which will help us improve the functionality, will be greatly appreciatedJ


Thursday, December 11, 2014

sys / as sysdba should require a password



sys / as sysdba should require a password

Applies to:
            Enterprise Linux – Version: 5

Description:
            Sqlplus / as sysdba connections go through OS authentication rather than database authentication. Any user with OS access can log into database irrespective of valid account available in database. Sys / as sysdba could not login without a password.
How can we control OS authentication to stop any OS user to login as SYSDBA?

Solution:
Set the following parameter in init.ora file:
REMOTE_OS_AUTHENT = FALSE

open sqlnet.ora file and set the following parameter:
SQLNET.AUTHENTICATION_SERVICES=(NONE)

save and exit file.

now you cannot login "sys / as sysdba" without password.


Your comments, especially which will help us improve the functionality, will be greatly appreciated :)

Tuesday, September 16, 2014

Auto Mount of Windows Shared Folder on Linux



Auto Mount of Windows Shared Folder on Linux

Applies to:
            Enterprise Linux – Version: any
           
Description:
            I have oracle database 11gR2 on oracle enterprise Linux 5.7. I want to mount a windows shared folder on linux.

Solution:
            To mount a windows shared folder on linux:
·         Login on linux by root user
·         Check the following rpm’s:
o   Fuse-2.7.4-8.el5
o   Fuse-ntfs-3g-2010.5.22-1.el5.rf
·         If not, install the above rpm’s on linux
·         Now create a new folder</data> on linux to mount windows folder
·         Grant read/write permission on windows shared folder<Data_Backups> to a windows/domain user<datacell>, windows machine IP <192.168.1.1>
·         Execute the following command on linux by root user to mount it:
mount  -t cifs  -o username=datacell,password=data123,ip=192.168.1.1 ‘\\192.168.1.1\Data_Backups’ /data
·         Windows shared folder mounted on /data on linux
·         Check mount points execute command:
df  -h
·         /data has been mounted

Now after restart linux machine mount point /data has been disappeared. To auto mount the /data mount point at startup on linux, you should enter the mount point entry in /etc/fstab as follow:

vi /etc/fstab

Go to bottom of  the file and enter the following line:

\\192.168.1.1\Data_Backups   /data     cifs      username=datacell,password=data123 0  0

Save and exit from file.

Now after restart the linux machine /data mount point automatically mount.


Your comments, especially which will help us improve the functionality, will be greatly appreciated :)

Thursday, August 7, 2014

ORA-00600: internal error code arguments: [kdsgrp1]


Applies to:
            Enterprise Linux – Version: 5
            Oracle Database 11gR2
           
Description:
             I have an update statement which modifies around 30000 records. Statement executed successfully on a single record, but when I execute to update all records it is going to fail with following error message:

ORA-00600: internal error code, arguments: [kdsgrp1], [], [], [], [], [], [], [], [], [], [], []

Solution:
           
·         Open and view the alert log file when error occurred
·         Get the trace file name from alert log file and open it
·         Search the keyword “Plan Table” in trace file
·         Here you found details of the query causing problem
·         Identify the tables from above query and execute the following:

Analyze table <tab_name> validate structure cascade;

·         Identify all the indexes on the above tables and execute the following:

Analyze index <index_name> validate structure;

·         Here I found an index which fail to validate and return the below error:

ORA-00600: internal error code, arguments: [kdsgrp1], [], [], [], [], [], [], [], [], [], [], []

·         Now I drop the index and recreate it.

·         After doing all the steps I resolved the above error.



Your comments, especially which will help us improve the functionality, will be greatly appreciatedJ

How to Extend Swap space on LVM Disk Linux

How to Extend Swap space on LVM Disk Linux   Applies to:             Oracle Database 12.2             Oracle Linux 7 Description: ...