Improve life-cycle management of web proxy server about rt-5gms-application-server

[davidjwbbc]

Have implemented process restarting logic (failure in event of rapid restarts). Have yet to implement the configuration reloading logic. This could be done either as:

1. Reload the AS configuration and ContentHostingConfiguration file, restart nginx.
2. Reload the AS configuration and ContentHostingConfiguration files if they've changed, restart nginx if any changes.
3. Reload the AS configuration and ContentHostingConfiguration files if they've changed, restart nginx if changes affect the nginx configuration.
4. Reload the AS configuration and ContentHostingConfiguration files if they've changed, choose restart of reload of nginx dependant on what has changed in nginx configuration.

[rjb1000]

After discussion, we decided that option 1 is too blunt.

I think for MVP#1, option 2 is good enough.

Options 3 and 4 are nicer, of course, but represent gold plating that can't be justified at this point.

[rjb1000]

As for the mechanism used the restart nginx: there is a brute force restart that unceremoniously terminates existing client connections and a graceful restart that hands over to a new process.

For the purposes of MVP#1, the brute force restart is good enough. We may revisit this later to make it more graceful.