@jamesthebard That's quite some hardware you got there. Do you happen to know the energy draw? And maybe have a few more insights documented?
Always curious how other people run stuff :)
Top-level
@jamesthebard That's quite some hardware you got there. Do you happen to know the energy draw? And maybe have a few more insights documented? Always curious how other people run stuff :) 3 comments
@jamesthebard If you run hwmon with node_exporter or alike, maybe you get power metrics from your CPUs: node_hwmon_power_watt Just in case you are curious :D @sheogorath I haven't setup proper monitoring like that yet...it's one of those things I've been meaning to get to, then never actually getting around to doing. Uptime monitoring? Yes. Proper backups? Definitely. The other types of monitoring such as free space, CPU/Mem utilization, etc...err...no. |
@sheogorath It doesn't trip the breaker at full load...unfortunately I don't have any power metrics for the lab so that about the best I can do in that regard. However, I tend to do a lot of encoding work so being able to spin up encoding nodes across the cluster is very nice. It also lets me try dumb/fun things like multipathing, overkill networking and the like.