Use node exporter on your machines and Prometheus to scrape the metrics and you will be drowning in pretty graphs. Super easy to set up and gives you system details across all your machines, all in one place.
There are also script exporters (a few implementations) for simpler Nagios style checks and more specialised ones for getting things like Cloudwatch data (shout out to YACE[1] - with dashboards from promcat). It's really quite vibrant the ecosystem.
Of course but it's still in the traditional domain of networking kit (generally). As there are many components virtualised or software defined then I think in those areas, not so much as there are alternative ways to get the metrics. SNMP is a bit long in the tooth and can be a bit unweildy sometimes, ymmv etc.
https://github.com/prometheus/node_exporter