Monitoring & Alerting

[nicolasberens]

We currently only collect data from systems via collectd or prometheus expoerter, but we should also be able to act on it.
There are many things that could be improved:

• For Servers we want to know about hardware failures. (via email)
• We should think about snmp querying switches.
• for nodes, an downtime notification might be interesting
  • maybe as a telegram bot where users can get notifications if their node is down for $X minutes

[FFHener]

SNMP querying for switches is currently started to be implemented in bbb-configs freifunk-berlin/bbb-configs/pull/739