Arbitrary Internal Topologies

[dastergon]

Currently, the TC Panel is based upon the Ansible Inventory which is a text file with fixed data that contains the topology information of the infrastructure. In case of decommissioning of a rack or in any event of a failure in the servers, TC Panel will continue to assume that the hosts are active. Unfortunately, Ansible is master-less and it does not keep state since it does not have agents installed on the hosts to notify the master server whether there is a failure. A simple solution would be to create simple health checks that are executed frequently by TC Panel in conjunction with the Ansible library. An advanced solution would the integration of a service discovery solution which can make TC Panel aware of the status of the systems.