goodrobots
Implement basic monitoring system
A basic monitoring system will be very useful:
- Monitor basic system parameters - CPU, disk space, memory etc
- Monitor basic flight parameters - dangerous config, battery use/level, inconsistencies (eg. one motor working too hard)
Currently what is used to obtain this info when populating the time series database? Can we interface that with the -api to allow the numbers to be viewable on a graphql driven dashboard?
collectd -> influxd. But this is quite a heavy solution both storage and memory/cpu-wise, so shouldn't be mandatory or a dependency on alerts/monitoring, imho. There are lots of opensource monitoring systems we can use. For example, checkmk (which is kind of an uber-wrapper around other monitoring systems like nagios) has a query language like system for monitoring status: https://checkmk.com/cms_livestatus.html That would be kind of cool I think?
Yeah, looks simple to interface with. Happy to roll with that as an option.