Write pgctl actions into the log

[chriskuehl]

Sometimes it's hard to tell when pgctl takes certain actions in the log. For example, it's not really clear at what point pgctl times out trying to start a service and then stops it (you can sometimes tell based on the finish script, but even then it's not clear why it exited: did pgctl kill it? did it die on its own?). It'd be useful if it wrote a line to the log like:

[pgctl] ERROR: service 'apache' failed to start after 120.12 seconds (it took 0.0468890666962s to poll), its status is up (pid 87214) 120 seconds

It already prints this, but doesn't put it into the log.

[chriskuehl]

Would also be nice to know at what point a service actually becomes ready, since the healthcheck may not print this.