Delete a Node from Oracle RAC 19c

 

Delete a Node from Oracle RAC 19c

In this Blog, we will see how to delete a node from multi node grid infrastructure setup. 

Current GRID NODES : Node 1, Node 2, Node 3 Node 

to be deleted : node3. Grid Version : 19c. 

STEPS OVERVIEW: 

1. Precheck. 

2.Delete Oracle Instance from node3 using DBCA. 

3.Deinstall the Oracle Database home from node3. 

4.Deinstall Grid Infrastructure home from node3. 

5.Delete clusterware config. 

6.Post Check.

1. Precheck :

Check whether all nodes are unpinned or not.

1	[oracle@node1 ~] $ olsnodes -s -t

Check the Cluster resources.

1	[oracle@node1 ~] $ crsctl stat res -t

2.  Delete oracle instance from node 3 using dbca. If any database instance is running on the node, then this need to be removed before deinstalling grid.

1	[oracle@node1 ~] $ crsctl status resource -w "TYPE = ora.database.type" -t

Run dbca from node1:

1 2	cd $ORACLE_HOME/bin ./dbca

Instance successfully deleted. Now proceed with deinstalling the oracle db home. 

3.  Deinstall the Oracle database home from node3.Specify the “-local” flag as not to remove more than just the local node’s software.

1	[node3]$ cd $ORACLE_HOME/deinstall -local

Few error like failed to delete directory got reported. We can delete those directories manually.

1	[oracle@node3]$ rm -rf /oracle/app/oracle

4.  Deinstall grid infrastructure home from node3.

1 2	[oracle@node2 ~]$cd $GRID_HOME [oracle@node2 ~]$./gridSetup.sh

RUN ON NODE3:

1	[root@node3 ~]# /crsapp/app/oracle/product/19c/crs/install/rootdeinstall.sh

5. Delete clusterware config.(Run on node 2) — as root user run below command from node 1

1 2	[root@node2 ~]# /crsapp/app/oracle/product/19c/bin/crsctl delete node -n DBHOST4 CRS-4661: Node DBHOST4 successfully deleted.

Delete remaining directories: Deinstall script might fail to delete few file/directories. So we can delete them manually.

1 2	rm -rf /crsapp/app/oracle/product/19c rm -rf /crsapp/app/oracle_base

6. Postcheck Run cluvfy postcheck script

1	[oracle@node2 ~]$ cluvfy stage -post nodedel -n DBHOST4 -verbose

Check updated cluster resource:

1	[oracle@node2s ~]$ crsctl stat res -t

Hope It Helps!!!!!