Post #1023601

Yesterday one of our load balancer servers went AWOL: it abruptly stopped responding to pings. Datacenter staff determined it had powered off and, also, it had a broken CPU cooler. I had a look at temperature graphs in Netdata, and indeed the poor thing had been running at 90°C+ for weeks. Netdata collects temperature metrics, but does not have pre-configured temperature alerts. I've now set up Netdata alerts to warn me about high CPU temperatures.