* update graphs -> systems page
* update availability graph to use the stats graph for 1 day averages
* a bit of cleanup for various graphs
* a few style fixes I missed
* another style fix
* update the pingperf graph as well
* update the poller perf graph as well
* style fix
* move graphs to using generic_stats default for readability
* Device Availability Calculation
* Travis fix
* .
* schema corrections
* flexible duration
* travis
* .
* .
* .
* .
* remove not needed code
* line Text to RRD
* update humantime
* .
* only set up again if device was marked as down
* set RRD area transparency
* save uptime also, to keep last availability as good as possible
* file description correction
* look for outages even if uptime > duration