Once a sysadmin….
After a 30 year career supporting IT systems through hardware, enterprise and (AWS) cloud, any sysadmin these days is mostly directed toward my home LAN and the Raspberry Pi that runs a few home-style bits and bobs for media/internet.
For what they are (and the cost) I think the rPi is a truly amazing device. Mine is a v3 (model B+) which runs our Plex media system, Deluge (for torrents), piHole DNS filtering (to block those annoying ads) and is also the SMB host for 3TB of USB3 connected network storage. It all runs happily with a Roku Express hooking in to Plex whilst covering Netflix + BBC iPlayer via solid WiFi streaming from a (recently installed) Vodafone hub and VDSL2 connection.
Having had a slight speed glitch from the new Vodafone solution I’ve been running a few manual tests to keep a check on things. You get more reliable results running any speed test from a hard wired device so I had been using Ookla’s speedtest-cli utility on the rPi. I realised it might be nice to automate this, maybe sending results to email, but other than setting up a cron task and forwarding output via ssmtp I wasn’t sure what might be the best (nicest) solution. My Brother suggested a visual output to a viewable webpage, which prompted me to look back into DevOps world and think about implementing a dashboard solution. Various companies I’d been involved with used Grafana extensively so it seemed like a good place to start.
Get richbos’s stories in your inbox
Join Medium for free to get updates from this writer.
The internet can be a useful place and I quickly dropped onto PiMyLifeUp who have a pre-made guide detailng exactly what I wanted to do. I’m no coder so wouldn’t have competency to put together the required Python logic, and although I’ve implemented and supported a few database clusters I’m not a natural DB admin, nor had I any experience with InfluxDB or any other type of time series database.
Thankfully the PiMyLifeUp guide is very comprehensive, and with pre-guides for both InfluxDB and Grafana I managed to get things running after only a couple of false starts (pre-running the python script to populate InfluxDB was key to configuring the Grafana data source).
I’ve set the cron task to 30 minutes, which would seem a reasonable amount of time to produce interesting enough graphs, although with no real fluctuation perhaps not that interesting…! Anyway, it was a nice little project to keep my hand in and inspired me to tinker further, setting up a Slack App to receive alerts should the speed drop.
So there we go, another small slice of usefullness added to the rPi 👍
