Restarting Standalone ASM instance post server bounce.

 

Below can be referred to in a scenario when you are required to starting Standalone ASM instance post server bounce, specially when the Database hosted on it is not coming up.

Step1: Check for Zombie processes.

ps -ef|grep d.bin

oracle   12945     1  0 Mar05 ?        00:02:08 /app/oracle/product/12.1.0.2/grid/bin/ohasd.bin reboot

oracle   13573     1  0 Mar05 ?        00:01:57 /app/oracle/product/12.1.0.2/grid/bin/cssdagent

oracle   13579     1  0 Mar05 ?        00:03:38 /app/oracle/product/12.1.0.2/grid/bin/oraagent.bin

oracle   13608     1  0 Mar05 ?        00:01:35 /app/oracle/product/12.1.0.2/grid/bin/evmd.bin

oracle   13621     1  0 Mar05 ?        00:01:39 /app/oracle/product/12.1.0.2/grid/bin/ocssd.bin

Step2: Kill all the process as they might be in hung state the same can be verified from CRS logs.

kill -9 13621   13608     

Step2: Login to root and stop and start OHASD service.

/etc/init.d/ohasd stop

/etc/init.d/ohasd start

Step3 (OPTIONAL): use crsctl command to check crs status to verify if it has come up automatically else start the same.

crsctl check has

crsctl check asm

 crsctl check css

crsctl check evm

if stopped

crsctl start has

crsctl start asm

crsctl start css

crsctl start evm

Step4 (OPTIONAL): check if oracleasm library has come up as a part of server bounce as this ensures external disks mapped to the server has also come up.

oracleasm status

oracleasm listdisks