scrutiny VS speedtest

I've noticed a quite nice feature in Scrutiny, where older disk temp data is lower resolution than recent data. In queries it would probably look like that:

I cannot look into the scrutiny script. The website recommends "something like"

I would always run the fantastic Scrutiny and set up notifications, so at the very least if some stat changes for the worst you might get a bit of early warning.

Hi. I am trying to understand whether I should trust TrueNAS Scale smartctl tool or Scrutiny (https://github.com/AnalogJ/scrutiny) for S.M.A.R.T results.

Quick Google search later to answer my own question, scrutiny seems to be it. That with LibreNMS sounds like a pretty healthy combo.

https://github.com/AnalogJ/scrutiny should do the trick

Its right there: https://github.com/AnalogJ/scrutiny/blob/master/docker/example.hubspoke.docker-compose.yml

I recently installed OpenWRT on a TP-Link TL-WDR4300, and put the router of my ISP in bridge mode. Now I noticed that the wireless speeds are significantly slower (40mbps vs 3mbps via librespeed.org), after using OpenWRT.

LebreSpeed

Try hosting a DIY speed test on a cloud server (like Google colab or the free oracle instances or whatever):

https://github.com/librespeed/speedtest

here's the top two results if you search for "open speed test nginx reverse proxy": https://github.com/librespeed/speedtest/wiki/Reverse-proxy-with-Nginx https://github.com/henrywhitaker3/Speedtest-Tracker/issues/924

Or https://librespeed.org/

Fast.com is giving me ~ 250Mbps https://librespeed.org is giving me ~ 112Mbps the one constant between all the tests seems to be the 52Mb upload

It should be DIA. They provide the internet connection to the company since 2 decades and it's a very small ISP, so it's very vague in terms of contract. Iperf was giving me very terrible results with TCP, UDP was giving me a couple of Gbit/s throughput, definitely a wrong result. We are using this self hosted speedtest. All my results above are based on this software: https://github.com/librespeed/speedtest

Put a copy of Librespeed on a web server that's accessible through the VPN and told them to use that. For (our) convenience, it's logged into a database that's correlated with the VPN login/logout times so the users don't even need to log in to use it, but we still know whose test result it is.