Logging utilities.

[Brendan Higgins]

Trying to revive this thread: we never concluded discussion on this. For the
convenience of the reader I have quoted the previous email below:

> Thank you for the background.  That explains a lot about the path Adriana
> has gone down.  I completely understand wanting to separate policy from
> reporting of device and process errors.  Where we differ is that I don't
> see that as the purpose of logging.  You've described an event-based
> telemetry system meant to be primarily consumed by other software to drive
> alerts and automated responses.  I view logging as being primarily for the
> benefit of a programmer or operator trying to understand what section of
> _code_ went wrong.  That distinction is part of why glog isn't designed to
> be super high-performance.  Google uses entirely separate systems for
> telemetry.  In my and Google's experience, scraping logs to extract
> information (even when the messages are well structured) winds up being
> painful to maintain and unreliable over time even when using journald.

We have a patch set out that adds glog support at:
https://gerrit.openbmc-project.xyz/#/c/788/

In addition, we have a patch for glog that adds support for logging to journald,
but I would like to get some feedback before I do anything else.