Deleting clusters when termination fails
Upon cluster termination, Data Hub only terminates the resources that it created. It does not terminate any resources (such as networks, subnets, roles, and so on) which existed prior to cluster creation.
The proper way to delete clusters is by using the Terminate option available from the cluster details in the Data Hub service console. If the terminate process fails, try the Terminate > Force terminate option. If the force termination does not delete all cluster resources, delete the resources manually:
To find the VMs, click on the links available in the Management Console > Data Hub Clusters > cluster details > Hardware section.
If you terminate instances on the cloud provider without using Data Hub, during the sync operation, Data Hub realizes that instances have been removed and marks the instances as DELETED_ON_PROVIDER_SIDE.
On AWS and GCP, Data Hub automatically removes all VM dependencies such as attached disks, network interfaces, public IPs and storage accounts. On Azure, you must manually delete the disks created for that instance and release the public IP assigned to that instance.
- To find the network and subnets, see the Cluster Information pane in the cluster details.