Lack of information output when a tang server is unreachable

[mralexc]

I ran into an issue today where the role fails at "Check whether devices are at the desired state" with no_log true. After disabling no_log, I get msg: 'Request failed: <urlopen error timed out>'

This makes sense in hindsight now that I have tracked down that 1 of several tang servers was unreachable. The error is very unhelpful and I think a check prior to this could be used to ensure availability of all tang servers could be added. This would produce helpful troubleshooting output, rather than a fail on a no_log.

[richm]

@mralexc The problem is that it will be tricky to output something to the log from this module since it deals with sensitive data. However, there is now a role variable nbde_client_secure_logging - the default is true - you can set this to false to turn off the no_log and get log data. I realize that there is still the issue that the role should give you a list of the tang servers with the connection issue with each one, if any.