What to choose to manage a distributed application?

Question

Clusters

I would like to find a solution for managing a distributed application on several cluster nodes. Each cluster node runs a separate application (service) instance.
All nodes run under Red Hat Linux.
Ideally, I would like to see the following functionality in the web admin panel:
1. Monitoring the load of the processor, memory, network interface on each node of the cluster
2. The ability to start / stop the application on all nodes of the cluster
3. Monitor the running of the application on all nodes
4. The ability to update applications on all nodes (for example, I upload tar with a new version of software to the web admin panel, which is distributed to all nodes)
5. The ability to launch an interactive console on a separate node and execute a couple of commands.
What is not needed:
- virtualization management (all nodes are iron servers)
And the most important condition: I would like the solution not to be based on containers.