add ability to trigger transcoder health-check from the CLI port

[iameli]

Orchestrators already the ability to self-check that they can transcode, and they do so on boot. We run into problems with "stuck Os" from time to time. If we add the ability to trigger the transcode self-check from the CLI port (perhaps POST http://localhost:7935/self-check), we can easily build a Kubernetes livenessProbe that will terminate stuck Os if they can no longer transcode.

[leszko]

Related to #2076.

In a more general sense, we want to be able to tell if Livepeer Node is up and ready, so that we can use it in k8 livenessProbe and readinessProbe.