ndt-server VS pihole-exporter

Data source: https://www.measurementlab.net

I get a test from https://www.measurementlab.net/ with it. Do realize that I'm from Europe and that testing a connection with an overseas server has a major impact on performance.

This is one of the better ones, but you can run whatever one you want, just don't use speedtest.net as they're trash now.

Tests are using the first hit when Googling "Speedtest" which connects to https://www.measurementlab.net/

To be specific in my thought here: we could fairly easily allow an admin to modify the endpoint in our client to measure loss & latency to an internal address or service in the cloud rather than our default. Speeds on the other hand are something we can't support just yet as it would require a test server near that endpoint. For what it's worth, the speed test application we use does have other examples for deploying your own test server. We'd need to do a deep dive into what that looks like from an engineering and business perspective before enabling teams to spin these up and use them as targets.

At a RIPE meeting I received this tip. Run an NDT test on https://www.measurementlab.net Although my skepticism at first I always seem to be getting a reliable response.

Google now uses https://www.measurementlab.net when you search for Speed Test so I have that bookmarked now

one possible solution: https://www.measurementlab.net/tests/ndt/

Well that is interesting, it's definitely something with googles built in speed test (thru https://www.measurementlab.net/).

I was just using speedtest.com and measurementlab.net/ before.

Using a Pi-hole to Prometheus exporter, this if I remember correctly: https://github.com/eko/pihole-exporter

This is definitely overkill, but it you already happen to be doing other monitoring of self hosted things there is a great Prometheus exporter for Pihole: https://github.com/eko/pihole-exporter.

DNS Ad blocking is from https://github.com/eko/pihole-exporter