Having fun with the Combustion Thermometer
The code for a Raspberry Pi application that connects to the Bluetooth thermometer, pulls data, and optionally uploads to S3 is in src/main.rs.
The bluetooth interface is the Rust program in this repo. It's responsible for talking to the thermometer and exposing the information to wherever it needs to go.
Simply cargo run to start the program scanning for the bluetooth accessory and print temperature updates.
To push data to an S3 bucket add the bucket name as a parameter: cargo run <BUCKETNAME>. It uses the AWS SDK so it will get credentials from the environment. The bucketname will be <YOUR NAME>-combustion as from the cdk below.
Run the raspberry pi interface to the ST7789 TFT by simply python3 display.py. It assumes the Rust program is running.
The Rust program uses the Bluer library which is the official Rust interface to the BlueZ Linux Bluetooth protocol stack.
I had to upgrade BlueZ including installing some dependencies.
Edit systemd/rustbustion.service and find the Enviroment line. Edit the AWS_PROFILE line to point to your profile name (stored in ~/.aws/credentials).
Copy systemd/rustbustion.service and systemd/display.service to /etc/systemd/system/, then
sudo systemctl daemon-reload
sudo systemctl start rustbustion.service
sudo systemctl start display.service
to start them, verify they work, then sudo systemctl enable <both services> to enable them to start at startup.
There is a simple implementation for MacOS that returns fake values for the purposes of testing.
This repo also contains a web application that will display the latest temperature data. The code is in src/bin/webapp.
For developing you can simply cargo run --bin webapp <bucket name> for running locally and open up http://127.0.0.1:8080.
To build and push the docker image you can:
$ ./scripts/build_docker_webapp.sh <name>
Where <name> is what you provided to the CDK stack (and also assumes that's your AWS credentials profile name).
Then follow the ECR instructions to push your locally tagged webapp:latest image to your ECR repository and update your service causing a deployment. You'll probably need to manually stop the previously running container because the t2.micro only has room for one container.
cdk/ contains CDK code to stand up a self-contained stack in AWS. It stands up:
- An S3 bucket for the Raspberry Pi program to upload to.
- An ECR repository to push a Docker image to.
- An autoscaling group with 1 t2.micro EC2 instance
- An ECS cluster that runs the webapp container on the EC2 instance.
cdk bootstrap --context name=<YOUR NAME> to bootstrap your AWS account.
cdk synth --context name=<YOUR NAME> to synthesize the template.
cdk deploy --context name=<YOUR NAME> to deploy the template. The S3 bucket is <YOUR NAME>-combustion which you can use for the Raspberry Pi program.