Monitoring means many things to many people and the important metrics are almost always something custom and specific, here is how monitor and alert on various bits of the lab.

Zabbix is my Environment monitoring, this monitors all physical hardware, interfaces (ILO, IMPI,IDRAC), power infrastructure (PDU through N.U.T. on a Raspberrry pi), Network Infrastructure via SNMP, the Proxmox hypervisor, Proxmox Backup Application, and Truenas all via API. Alerting is configured for e-mail alerts.

The Proxmox hypervisor has built-in alerting that notifies me on both some hardware issues through the hypervisor but critically on automation and backup jobs that run.

Applications are monitored one of two ways, if I am monitoring for the speed of a certain metric then ful log/diagnostic data is brought in to prometheus/Grafana for analysis. I do not do any alerting out of this.

For basic uptimes and response times I use UptimeKuma, it's a very simple tool to notify when sites go down and keep some metrics for the data. This could also be done in zabbix but I provide this monitoring service to several friends and wanted to keep it simple.

Network traffic is logged and visualized in Akvorado, no alerting is configured.