More strict health check for providers

[Gohlub]

   async fn health_ping(&self, arg: DummyArgument) -> Result<String, String> {
        info!("Health ping received: {:?}", arg);
        Ok("Ack".to_string())
    }

this check only makes sure that the node is alive, not that its actively serving the provider that the operator might be searching for. In the provider code, we actually only check whether the provider exists after getting the tx hash, which is bad. The health check should also include the providerID so that we can programmatically verify that at least a definition of that provider exists before sending payment.

We both need to make changes for this to happen, I can wait on you to update the health check before I make my update to understand what structure to expect. I would say that this is pretty high on our priority list (as it can enable faulty irreversible payments

[Hallmane]

Right, and also if they aren't responding correctly, then updating the registry, marking them as offline or something towards that direction.