PiMonitor

A CPU utilization monitoring service for the Raspberry Pi 3 Model B implemented using the ClearBlade SDK and ClearBlade Platform.

Quickstart

Monitor

First, on the ClearBlade platform, create a new System. Then, create a Device within this system. Next, clone this repository onto the Raspberry Pi and create a config file in the root directory of the project named clearblade.ini. Fill in the details here with the information from the system and device you just created. Feel free to choose any channel name you like.

[system]
key = <system key>
secret = <system secret>

[device]
name = <device name>
active_key = <device active key>

[messaging]
channel = <channel name>

Service

Create a new Service on the ClearBlade platform in the System you created earlier. Add the following settings to this Service:

• Security: add the Authenticated role and allow it to execute the Service
• Requires: add the clearblade requirement
• Triggers: add a new trigger with source as Messaging, action as Publish, and topic equal to the channel name you set in the above configuration file.

Next, paste service/cpuCollectionService.js into the code box on the Service page on the ClearBlade platform site.

Collection

Create a new Collection named rpiCPU in the same System you've been working in. Add the following columns:

• name: cpu_0, type: float
• name: cpu_1, type: float
• name: cpu_2, type: float
• name: cpu_3, type: float
• name: time, type: timestamp

Monitor, part 2

Finally, install and run the monitor.

# create a virtual environment
python3 -m venv venv
. venv/bin/activate
# install dependencies
pip3 install -r requirements.txt
# run monitor
python3 run.py

You should see rows start to show up on the Collection page on the ClearBlade platform. A message will be sent every 5 seconds to the ClearBlade platform with the percentage usage of each core of your cpus.

Testing

python3 run_tests.py

Assumptions

This program is tailored for use on a Raspberry Pi 3 Model B. As such, this program assumes that the CPU of the machine it is running on has 4 (and only 4) cores.