You disable high availability for Db2® Big SQL on HDP in the Ambari dashboard. Disabling HA is the task of
decommissioning and deleting the standby head node. On CDH, use the Db2 Big SQL cluster administration utility to disable high
availability for Db2 Big SQL.
Before you begin
Because the Db2 Big SQL home directory is removed during
the decommissioning of the Db2 Big SQL standby head node, be
sure to back up any private data stored in Db2 Big SQL directories on the standby head before you
modify the HA configuration.
Procedure
- Check if both the primary head node and the standby head node are running.
- CASE 1: If both the primary head node and the standby head node are running, follow
the steps below:
- Log into Ambari as an administrator and select the IBM®
Db2 Big SQL service.
- Under the Summary tab, select the standby (inactive) head node. Then, in the Db2 Big SQL head, select the Started drop-down menu and click
Turn On Maintenance Mode.
- While the node is in maintenance mode, select the Decommission
option from the Started drop-down menu.
- From the Stopped drop-down menu, select
Delete to delete the head node component from the cluster.
- CASE 2: If the standby head node is down or cannot be brought up, complete the
following manual steps to disable the inactive standby head node.
Note: These steps must be completed as the root user or under the non-root install user ID.
Assume that the host that contains the primary head node is
primaryheadnode.com and that the host that contains the standby head node is
standbyheadnode.com.
-
Run the following command, first on the standby head node, and then on the primary head
node:
bigsql stop -failoverController
-
Remove the Db2 Big SQL head node from the standby head
host that is down:
curl -u admin:admin -H "X-Requested-By: ambari" -X DELETE
http://server-name:server-port/api/v1/clusters/cluster-name/hosts/standbyheadnode.com/host_components/BIGSQL_HEAD
- Run the high availability cleanup script on the primary head node:
/var/lib/ambari-agent/cache/extensions/IBM-Big_SQL/5.0.4.0/services/BIGSQL/package/scripts/bigsql-HA-cleanup.sh
-P primaryheadnode -S standbyheadnode -U bigsql
-
Remove Db2 Big SQL from the faulty standby
head node.
Note: Be sure to run this command on the faulty standby head node. Running this
command on the primary head node causes Db2 Big SQL to be
uninstalled.
/var/lib/ambari-agent/cache/extensions/IBM-Big_SQL/5.0.4.0/services/BIGSQL/package/scripts/bigsql-drop-host.sh
-p /usr/ibmpacks/current/bigsql -n standby_head_node_host_name
-u bigsql -d 'comma-separated_list_of_data_directories_and_the_database_directory'
-v /var/ibm/bigsql -G Db2_admin_group_name -c True -t 0
For
example:
/var/lib/ambari-agent/cache/extensions/IBM-Big_SQL/5.0.4.0/services/BIGSQL/package/scripts/bigsql-drop-host.sh
-p /usr/ibmpacks/current/bigsql -n headnode2.ibm.com
-u bigsql -d '/var/ibm/bigsql/database,/hadoop/bigsql'
-v /var/ibm/bigsql -G bigsqladm -c True -t 0
-
After this cleanup, you can set up high availability by using a different node as the Db2 Big SQL head.