Oracle Exadata Database Machine Shutdown and Reboot Procedure

This example uses a quarter rack X2-2 for an example.  The quarter rack composes of 2 database node and 3 cell node.

To Shutdown
Go to DB Node 1

1. Login as root
2. use command . oraenv  to change the Oracle base to +ASM1 instance
3. type crsctl status res -t | more to check the Oracle service status
4. type ps -ef | grep pmon to check oracle instances
5. which crsctl to check the oracle home.  We should use the crsctl under /u01/app/11.2/grid/bin/crsctl
6. shutdown Oracle: crsctl stop crs
   You should see the message "CRS-4133: Oracle High Availability Services has been stopped." as the last message.
7. Double check if all Oracle instances have been closed:  type ps -ef | grep pmon
8. shutdown the Linux: shutdown -h -y now

Go to DB Node 2

1. Login as root
2. use command . oraenv  to change the Oracle base to +ASM2 instance
3. Repeat all other steps just like those on Node 1

Go to Cell Node 1

1. login as root
2. use the command service celld status to check the status of storage service
   You should see rsStatus, msStatus and cellsrvStatus are "running".
3. stop the service: service celld stop
4. We can use service celld status to double check if rsStatus is stopped.
5. Shutdown the Linux:  shutdown -h -y now

For Cell Node 2 and 3, repeat the same steps as Cell Node 1

To power on

1. Power on Cell Node 1,2 and 3
2. Double check if the Cell service is up by using command  service celld status
3. Then power on DB Node 1 and Node 2.
4. Double check if DB status is normal:  crsctl status res -t | more