@Frankenscript , @bp2008
Sure, hopefully this helps. Refer to: ESP8266 Publishing DHT Readings Raspberry Pi | Random Nerd Tutorials for a basic idea that I started from. My end point was a little different.
Sure, hopefully this helps. Refer to: ESP8266 Publishing DHT Readings Raspberry Pi | Random Nerd Tutorials for a basic idea that I started from. My end point was a little different.
- ESP8266 running "ESP Easy Mega" firmware.
- This firmware sure beats trying to write your own program for the ESP8266, and has built in web server for configuration.
- I originally bought from a guy on Tindie that has been great to work with, and even helped me troubleshoot because it came as a "kit" in a 3D printed case for WAF. BK Hobby on Tindie
- this kit has an OLED display and a DHT22 temperature & humidity sensor that has been reasonably accurate. "ESP Easy" allows you to set offsets for humidity and temperature if you have a good reference to calibrate from.
- ESP8266 only connects via wifi, but it's just sending little text messages (basically) to the MQTT server (every minute). I have it auto-connect to the IOT wifi network on my Ubiquity network.
- "ESP Easy" firmware provides support for formatting and sending MQTT to an "Open HAB MQTT" (controller they call it), and you can configure the format (which is important later when breaking it down and inserting it in node-red).
- Each Temperature sensor sends the temperature and humidity every 60 seconds (possibly overkill), this can be configured right in ESP Easy.
- A firewall rule so the IOT devices can send MQTT messages only to the Raspberry Pi (via port config).
- Raspberry PI 3B+ (because that's what was available at the time)
- Running basic Raspian because I don't know any different
- running node-red. Currently node-red just grabs the incoming MQTT, builds a SQL INSERT statement and fires it off to the MS-SQL server for storage.
- running Mosquitto MQTT server
- A custom firewall rule so the Raspberry Pi (on the IOT subnet) can talk to the MS-SQL machine on a different (video) subnet.
- Blue Iris machine (always on), running Microsoft SQL Server Express
- Because it was always on anyhow.
- Running Microsoft SQL Express because it was free and I happen to know my way around that database, MySQL would have been fine as well
- I probably need to lower the frequency I send the temperature readings or summarize the historical data. I currently have 330 MB SQL file for past 1.5 years (4,464,923 data points).
- Tableau (or Grafana if going the free route, with a little Python)
- Because I have it, and I just haven't found the time to get Grafana working for visualization on Raspberry Pi. Obviously, this would be overkill for a purely DIY option, Grafana is probably the best documented option.
Last edited: