Remove broken configtest_enable option

[wyardley]

Per #916, the $configtest_enable option doesn't seem if it could possibly work. In general, it would be good to review if everything here does what it is supposed to, but I think this will do the right thing on most platforms without any additional work.

I'm not sure if anyone would be using the $service_restart command directly, I'd think in most cases, we should be relying on $service_name and letting Puppet do the right thing, though I'm open to suggestion if people think $service_restart still needs to be exposed directly. What I don't understand is that, though it defaults to /etc/init.d/nginx reload, which doesn't work, the systems I tested all were using systemctl to restart the service anyway.

We still may want to revisit some kind of reporting back if the config test fails in a more generic way. I'm not going to address that one way or another for now.

[wyardley]

passed acceptance tests for
centos-66-x64
centos-72-x64
ubuntu-server-1404-x64
debian-78-x64

Getting a (presumably unrelated) error running on debian-82-x64

/usr/local/lib/ruby/gems/2.3.0/gems/beaker-2.51.0/lib/beaker/host.rb:351:in `exec': Host 'debian-82-x64' exited with 1 running: (Beaker::Host::CommandFailure)
 puppet agent --configprint hiera_config
Last 10 lines of output were:
        from /usr/lib/ruby/vendor_ruby/puppet/util.rb:14:in `<top (required)>'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/lib/ruby/vendor_ruby/puppet.rb:8:in `<top (required)>'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/lib/ruby/vendor_ruby/puppet/util/command_line.rb:12:in `<top (required)>'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/lib/ruby/2.1.0/rubygems/core_ext/kernel_require.rb:55:in `require'
        from /usr/bin/puppet:7:in `<main>'
    from /usr/local/lib/ruby/gems/2.3.0/gems/beaker-2.51.0/lib/beaker/host.rb:39:in `[]'
    from /usr/local/lib/ruby/gems/2.3.0/gems/beaker-2.51.0/lib/beaker/dsl/install_utils/foss_utils.rb:300:in `block in install_puppet_on'

[tux-o-matic]

@wyardley you removed a feature by deleting 'service_restart'.
One example, Docker, when running Nginx within a container rather than a initd/systemd service, you can use 'service_restart' to have the Puppet module interact with the container.

[wyardley]

@tux-o-matic: I can re-implement $service_restart separately (though I'd argue that it's generally a Good Thing to just use the $service_name parameter most of the time, and let puppet handle the service itself).

In my earlier test, even with $configtest_enable set, I couldn't get Puppet to call the specified custom init script, however, I am able to get that to work now. It is easy enough to restore this functionality without $configtest_enable; I'll open another PR for that (#927)

[tux-o-matic]

Thanks @wyardley. For Nginx users wishing to mix Puppet and Docker, you can't just rely on the Puppet Service type (until we get a Docker provider for Service).