We currently use ansible_fqdn in the playbook that automates the addition of nodes to PMM. This poses a few issues:
- the hostname may have no resemblance to the CNS entry -> confusion
- the hostname cannot be resolved by ansible -> localhost.localdomain -> confusion
- the customer wants the host labelled differently
Currently there is only the option to remove and add back the node with a different name, thus losing visibility of the already accumulated data.
Adding pmm-admin rename <existing name> <new name> or similar would allow the Consul and QAN data to be "fixed". Loss of Prometheus data would need to be accepted in such a use-case
Before renaming : re-check if User ready to rename and lost previous data on Graph (Y/n) and --force-rename option to rename without asking confirmation