Pages

Sunday, April 5, 2015

Steps to bring down and start RAC cluster

Use following steps to bring down Node  | cluster-11gr2

a) olsnodes -n
(gives number of nodes in a cluster)


b)
crsctl status resource -t

 (store output of the command)

c) stop agents on all nodes in a cluster  and create blackout in OEM as per requested timings
 emctl stop dbconsole
( stop the agent running on oracle home)

d) Perform following steps in all nodes in a cluster
 stop local listener and check status if listener
 

srvctl stop listener -n nodename
srvctl status listener -n nodename

e) Perform on any one node  

check how many databases running on RAC- CLUSTER?
bring down the all RAC databases

srvctl status database -d dbname(unique)


(This command shows, how many instances the databases running on RAC CLUSTER)
srvctl stop database -d dbname(unique)
srvctl status database -d dbname(unique)

f) Shut down asm running on each node in cluster with following command
srvctl stop asm -n nodename

check status if asm after
srvctl status asm -n nodename

g) stop node application running on the each node in a cluster
if you don’t do this if HAS is enabled, it automatically reboot the nodes
make sure stop node apps on all nodes

srvctl stop nodeapps -n nodename

h) Make sure these steps performed on all nodes in a cluster from any of the node issue following command and stop cluster-ware stack on RAC 
crsctl stop cluster -all
( -all bring the cluster-ware stack on all nodes in a cluster)

after this command is completed.
make sure CRS Stack is down on all nodes with following command


crsctl check cluster -all
crsctl_stat -t


######------Your cluster is down ------######
Use following commands to bring up RAC Cluster

a) Maintainance

(Following command brings up CRS stack in all nodes in a cluster)
crsctl start cluster -all

check the status of CRS in all nodes in a cluster using following command 
crsctl check cluster -all

b) start node applications on each in a cluster
srvctl start nodeapps -n nodename

check status on each node
srvctl status nodeapps -n nodename

c) Startup asm running on each node with following command
srvctl start asm -n nodename

check status if asm after
srvctl status asm -n nodename

d)bring up all databases in a cluster
srvctl start database -d dbname(unique)

check the status of database with following command
srvctl status database -d dbname(unique)

e)Start Listener running on all databases in a cluster
srvctl start listener -n nodename

check status if local listener on each node in a RAC cluster

srvctl status listener -n nodename

f) start agents on all nodes 

srvctl start dbconsole

validate all components with following command, check status of each resource on a cluster with following command and compare the result set with the before 


crsctl status resource -t


---- NIKHIL TATINENI ----

---- Oracle In Memory ----