To remove an existing cluster node, use the AWS Management Console and perform the following steps.
Set the value of the tag "fujitsu.pclswr.is_recovery_target" of the instance of the cluster node to false.
Stop the instance of the cluster node.
Note the integer value and instance ID that identifies the cluster node with the tag "fujitsu.pclswr.id" for the instance that you stopped in step 2.
Terminate the instance that you stopped in step 2.
Deregister the AMI for the instance that you stopped in step 2.
If you created a snapshot on a volume attached to the instance you stopped in step 2, delete the snapshot.
Delete the load balancer for the ELB used by the instance in step 2.
Delete the ELB target group used by the instance in step 2.
Delete the mounted EFS file system on the instance in step 2.
Delete the CloudWatch alarm that you configured for the instance in step 2. Determine what to delete by the integer value that identifies the cluster node at the end of the alarm name. For more information about alarm names, see "A.2.4.3 Configuring CloudWatch Alarm".
Remove the security group that allows the instances in step 2 that you are configuring on your VPC endpoint.
Delete the integer value that identifies the cluster node that you want to delete from the AWS Lambda environment variable "PCLSWR_SYSTEM_LIST". For more information about environment variables, see "A.2.4.1 Configuring AWS Lambda".
Remove the integer value that identifies the cluster node to remove from the subnet tag "fujitsu.pclswr.idlist". For more information about tags, see "A.2.1.1 Creating VPC and Subnet".
Delete the item for the deleted instance that is registered to the table "Fujitsu-Pclswr-DB-Switcher" in Amazon DynamoDB. The items for the instance to be deleted are checked from the items "SystemID" and "InstanceID". For more information about the items, see "A.2.4.4 Configuring Amazon DynamoDB".
