Add a test for bluechictl monitor <node> <service>

[mwperina]

• Flush stdout in bluechictl monitor after each output to stdout
• Add a test for bluechictl monitor

Fixes: #777

[coveralls]

coverage: 77.871% (+1.6%) from 76.26%
when pulling 76da056 on mwperina:test-bluechictl-monitor
into 68d893c on eclipse-bluechi:main.

[mkemel]

I was thinking about another solution for that, that would also test bluechictl metrics listen
Add a class in bluechictl.py that would run bluechictl in another thread, having start and stop methods, with stop returning the output, which you can check for anticipated output. That would be a more general and reusable solution. WDYT?

[mwperina]

Switched to use temporary file for parsing bluechictl monitor results, so we cleanup properly after the test finishes

[engelmi]

I was thinking about another solution for that, that would also test bluechictl metrics listen Add a class in bluechictl.py that would run bluechictl in another thread, having start and stop methods, with stop returning the output, which you can check for anticipated output. That would be a more general and reusable solution. WDYT?

In general, there is probably a lot of optimization potential in the current tests - this could be one. Lets focus on the 100% function coverage first and optimize after that :)