Home » Server Options » RAC & Failsafe » Restart RAC DB (Oracle 11.2.0.3.0 , Unix)
Restart RAC DB [message #577433] Fri, 15 February 2013 09:55 Go to next message
Mohan10g
Messages: 159
Registered: May 2009
Location: INDIA
Senior Member

Hi,

I want to set the following parameter in one node and restart the machine.
ALTER SYSTEM SET MEMORY_MAX_TARGET=4G SCOPE=SPFILE sid='*'

I know we could use srvctl command to restart DB.My question is..to restart DB is it advisable to use
shutdown immediate and startup command ?

Thank you
Re: Restart RAC DB [message #577435 is a reply to message #577433] Fri, 15 February 2013 10:08 Go to previous messageGo to next message
John Watson
Messages: 4870
Registered: January 2010
Location: Global Village
Senior Member
I don't understand your question. Do you mean this -
srvctl stop instance -d dbuniquename -i instancename -o immediate
srvctl start instance -d dbuniquename -i instancename -o open
If so, those are the default options.
--
John Watson
Oracle Certified Master
http://skillbuilders.com
Re: Restart RAC DB [message #577455 is a reply to message #577435] Fri, 15 February 2013 22:03 Go to previous messageGo to next message
Mohan10g
Messages: 159
Registered: May 2009
Location: INDIA
Senior Member

Hi John,

My question is can we use shutdown immediate and startup command to restart DB after setting the above ALTER SYSTEM command ?

I hope my question is clear now.

Re: Restart RAC DB [message #577457 is a reply to message #577455] Sat, 16 February 2013 00:56 Go to previous messageGo to next message
Michel Cadot
Messages: 60060
Registered: March 2007
Location: Nanterre, France, http://...
Senior Member
Account Moderator
You can but it not the proper way.
The proper way is to use srvctl.
Do you understand what is RAC and its architecture?

Regards
Michel
Re: Restart RAC DB [message #577458 is a reply to message #577455] Sat, 16 February 2013 02:15 Go to previous messageGo to next message
John Watson
Messages: 4870
Registered: January 2010
Location: Global Village
Senior Member
As I said, SHUTDOWN IMMEDIATE is what srvctl will do by default. If you prefer to save time with a SHUTDOWN ABORT, then use the option
-o abort
It is all in the help, you know:
srvctl -h

Rather annoyingly, srvctl does not have a FORCE option, so if you want to user STARTUP FORCE you need to do that with SQL*Plus.
--
Update: typo, missed out "not" in my final sentence. Corrected now.

[Updated on: Sat, 16 February 2013 02:16]

Report message to a moderator

Re: Restart RAC DB [message #577461 is a reply to message #577458] Sat, 16 February 2013 03:17 Go to previous messageGo to next message
Michel Cadot
Messages: 60060
Registered: March 2007
Location: Nanterre, France, http://...
Senior Member
Account Moderator
Quote:
if you want to use STARTUP FORCE you need to do that with SQL*Plus.


Do NOT use this statement but in emergency case although I don't see the advantage of this one above SHUTDOWN ABORT (which also should be used only in emergency case) + STARTUP.
Too many people use STARTUP FORCE for anything, just to avoid typing 2 commands. It is a bad habit.

Regards
Michel

[Updated on: Sat, 16 February 2013 03:19]

Report message to a moderator

Re: Restart RAC DB [message #579516 is a reply to message #577461] Tue, 12 March 2013 18:07 Go to previous message
alan.kendall@nfl.com
Messages: 163
Registered: June 2012
Location: Culver City, California
Senior Member
Here are the commands that I use to restart RAC instances, ASM, CRS and Listeners.

Whenever I do maintenance on RAC, I stop then start one instance at a time. This way if
there is a startup error, I can fix it from one of the started instances and "Voila" I never
have an outage related to changing oracle memory parameters.
SYS AS SYSDBA> alter system set sessions=10000 scope=spfile sid='*';
System altered.
SYS AS SYSDBA> alter system set processes=10000 scope=spfile sid='*';
System altered.
SYS AS SYSDBA> alter system set open_cursors=10000 scope=spfile sid='*';
System altered.

>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe
>srvctl stop instance -d CSCDAP -i CSCDAP1
>srvctl status database -d CSCDAP
Instance CSCDAP1 is not running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe
>srvctl start instance -d CSCDAP -i CSCDAP1
>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe
>srvctl stop instance -d CSCDAP -i CSCDAP2
>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is not running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe
>srvctl start instance -d CSCDAP -i CSCDAP2
>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe
>srvctl stop instance -d CSCDAP -i CSCDAP3
>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is not running on node csprdcdadb03-fe
>srvctl start instance -d CSCDAP -i CSCDAP3
>srvctl status database -d CSCDAP
Instance CSCDAP1 is running on node csprdcdadb01-fe
Instance CSCDAP2 is running on node csprdcdadb02-fe
Instance CSCDAP3 is running on node csprdcdadb03-fe

Here are commands that I used to check and stop "ASM" before some maintenance.
The asm instances came up on their own when the Unix boxes were restarted.
csdrdb01(oracle)/u01/home/oracle>srvctl status asm -n csdrdb01
ASM instance +ASM1 is running on node csdrdb01.
csdrdb01(oracle)/u01/home/oracle>srvctl status asm -n csdrdb02
ASM instance +ASM2 is running on node csdrdb02.
csdrdb01(oracle)/u01/home/oracle>srvctl stop asm -n csdrdb01
csdrdb01(oracle)/u01/home/oracle>srvctl stop asm -n csdrdb02
csdrdb01(oracle)/u01/home/oracle>srvctl status asm -n csdrdb01
ASM instance +ASM1 is not running on node csdrdb01.
csdrdb01(oracle)/u01/home/oracle>srvctl status asm -n csdrdb02
ASM instance +ASM2 is not running on node csdrdb02.

To shutdown the database prior to shutting down ASM I use the following commands.
>srvctl status database -d CSCDAP
>srvctl stop instance -d CSCDAP -i CSCDAP1
>srvctl stop instance -d CSCDAP -i CSCDAP2
>srvctl stop instance -d CSCDAP -i CSCDAP3
>srvctl status database -d CSCDAP

To stop CRS I did the following:
cd /u01/app/oracle/product/10.2.0/crs
crsctl check crs
crsctl stop crs

To restart the listener on one node:
csstgcdadb11(oracle)/home/oracle>srvctl status listener -n csstgcdadb11
Listener LISTENER is enabled on node(s): csstgcdadb11
Listener LISTENER is running on node(s): csstgcdadb11
csstgcdadb11(oracle)/home/oracle>srvctl status listener -n csstgcdadb12
Listener LISTENER is enabled on node(s): csstgcdadb12
Listener LISTENER is running on node(s): csstgcdadb12
csstgcdadb11(oracle)/home/oracle>srvctl stop listener -n csstgcdadb11
csstgcdadb11(oracle)/home/oracle>srvctl status listener -n csstgcdadb11
Listener LISTENER is enabled on node(s): csstgcdadb11
Listener LISTENER is not running on node(s): csstgcdadb11
csstgcdadb11(oracle)/home/oracle>srvctl start listener -n csstgcdadb11
csstgcdadb11(oracle)/home/oracle>srvctl status listener -n csstgcdadb11
Listener LISTENER is enabled on node(s): csstgcdadb11
Listener LISTENER is running on node(s): csstgcdadb11
Previous Topic: Master node in oracle 9i rac
Next Topic: RAC Configuration
Goto Forum:
  


Current Time: Fri Dec 26 02:45:05 CST 2014

Total time taken to generate the page: 0.07470 seconds