RasPi `monitoring.fritz.box`: Unstable setup / insufficient resources

[sebastian-sommerfeld-io]

Probability	Impact
Medium	Medium

Description

Running Prometheus, Grafana, and Loki on a single Raspberry Pi node (monitoring.fritz.box) might carry the risk that the machine may become overwhelmed and experience unexpected and/or repeated crashes due to insufficient resources.

Affected services

• monitoring.fritz.box
• supervisor.fritz.box

Response

A temporary supervisor node has been installed to monitor the stability of the primary monitoring node.

If unexpected or repeated crashes occur, the solution is to split Prometheus, Grafana, and Loki and move each service to a dedicated Raspberry Pi node. By dedicating separate Raspberry Pi nodes to each service, the load can be distributed more evenly, reducing the strain on each machine and improving overall system stability.

[sebastian-sommerfeld-io]

Maybe stability issues are not related to the RasPi resources. Maybe they are related to network issues (with a node being unreachable from time to time).

[sebastian-sommerfeld-io]

RasPi is not in real use. All RasPi nodes need a new setup