ORACLE SWITCHING FROM DEDICATED SERVER TO SHARED SERVER (MTS)
By Edward Stoever
- Page 3
-
-
oracleasm createdisk failed
Solution: The ASM error "disk does not exist or is not instantiated" error can be caused by many issues:
- Not disabling SELINUX:
vi /etc/sysconfig/selinux SELINUX=disabled
Then reboot your system and run getenforce to confirm that
SELINUX is disabled:
root> /usr/sbin/getenforce Disabled -
Oracle 11G - Install oracle grid infrastucture 11GR2
-
TUTO Video - Installer ASM
-
Oracle 11G - réduire la taille du tablespacec temporaire
SQL> alter tablespace TEMP shrink space keep 1G;
Source. -
Réinitialiser le tablespace temporaire
-
Configuration E/S - Oracle
Oracle® Database Performance Tuning Guide
11g Release 2 (11.2)
Part Number E16638-03 -
Raw device pour Oracle
Using Raw Devices for Oracle Databases
Configure a raw device
-
Configure the desired device in the
/etc/sysconfig/rawdevices
file:/dev/raw/raw1 /dev/sdb1
-
Start and enable the device by running:
# service rawdevices start # chkconfig rawdevices on
At this point, you could set ownership and permissions withchmod
andchown
. However, because these devices will be created dynamically at boot time, permissions set in this manner will not persist after a reboot.Binding Logical Volumes to the raw devices:I have added the below lines into the /etc/sysconfig/rawdevices and restarted the rawdevices service to bind logical volumescreated in the above steps.
/usr/bin/raw /dev/raw/raw11 oracle_vg/lvol0 # test db -- SYSAUX Tablespace
/usr/bin/raw /dev/raw/raw12 oracle_vg/lvol1 # test db -- UNDOTBS1
/usr/bin/raw /dev/raw/raw13 oracle_vg/lvol2 # test db -- TEMP tablespace
/usr/bin/raw /dev/raw/raw14 oracle_vg/lvol3 # test db -- EXAMPLE tablespace
/usr/bin/raw /dev/raw/raw15 oracle_vg/lvol4 # test db -- USERS tablespace
/usr/bin/raw /dev/raw/raw16 oracle_vg/lvol5 # test db -- redo1_1.log
/usr/bin/raw /dev/raw/raw17 oracle_vg/lvol6 # test db -- redo1_2.log
/usr/bin/raw /dev/raw/raw18 oracle_vg/lvol7 # test db -- redo1_3.log
/usr/bin/raw /dev/raw/raw19 oracle_vg/lvol8 # test db -- control1
/usr/bin/raw /dev/raw/raw20 oracle_vg/lvol9 # test db -- control2
/usr/bin/raw /dev/raw/raw21 oracle_vg/lvol10 # test db -- control2
/usr/bin/raw /dev/raw/raw22 oracle_vg/lvol11 # test db -- spfile
/usr/bin/raw /dev/raw/raw23 oracle_vg/lvol12 # test db -- pwfile
/usr/bin/raw /dev/raw/raw24 oracle_vg/lvol13 # test db -- SYSTEM TablespaceAlso, you need to change the ownership of these devices to oracle user. Raw devices are refreshed with the default
permissions and ownership every time you reboot your system. For this reason, I add these lines to the /etc/rc.local so
that every time machine reboots, this devices are assigned correct ownership/permissionsPlease add the below lines
to the /etc/rc.localfor i in `seq 1 25`
do
chown oracle:dba /dev/raw/raw$i
chmod 600 /dev/raw/raw$i
done -
-
Installation ASM et configuration de l'instance +ASM
-
Quand utiliser le synchronuous I/O
En anglais :
You would be keen to disable asynchronous I/O when you see high average_wait on event db_file_parallel_wait. Other reason for turning it off will be synchronous I/O is more reliable.
SQL> select event,average_wait from v$system_event where event like 'db file parallel write';
EVENT AVERAGE_WAIT
---------------------------------------------------------------- ------------
db file parallel write 28.2 [ centi seconds]This is not a very good ASYNCH I/O. Try Synchronous I/O
Other parameters to affect write (as well as read) is dbwriter_processes. When asynchronous I/O operations are slower in operating system in comparison to synchronous I/O then turn off asynchronous I/O by setting disk_asynch_io to false and set multiple db writer processes by increasing dbwriter_processes values from 1 to 2,3 or 4 suitable value to your system. Alternate is incrase dbwr_io_slaves from 0 to 2,3,4 suitable value.
Note 1: Asynchronous I/O operations are more prone to block corruptions than synchronous operations so many DBAs disable it and follow practice as mentioned in above paragraph. So if you do not have standby databases and oracle 11g then which autoamatically recovers corrupted block on primary then you would not want asynchronous I/O
Note 2: For 11g R2 for tuning purpose, the “db file async I/O submit” should be treated as “db file parallel write” in previous releases.