为 Raspberry Pi 构建 Python IoTEdge 模块

Question

I am trying to build a Python IoTEdge module for a Raspberry Pi 4 (uses Docker to build image).

I set the target architecture for my IoTEdge solution as arm32v7

I don't believe anything is wrong with my requirements.txt , but just for reference:

requirements.txt

azure-iot-device~=2.0.0
opencv-python
numpy
requests

Dockerfile

FROM arm32v7/python:3.7-slim-buster

WORKDIR /app

COPY requirements.txt ./
RUN pip install -r requirements.txt

COPY . .

CMD [ "python3", "-u", "./main.py" ]

I run the following command:

docker build  --rm -f "path/to/Dockerfile" -t dockerTag "path/to/module/project"

However, I get the following error:

Step 4/7 : RUN pip install -r requirements.txt
 ---> Running in 515aac4bdc38
standard_init_linux.go:219: exec user process caused: exec format error

I read here that to build to ARM, I have to use docker buildx instead. I created a buildx container with --platform linux/arm32 . I switched to using this container, and run:

docker buildx build --platform linux/arm32 --rm -f "path/to/Dockerfile" -t dockerTag "path/to/module/project"

However, now I get:

 => ERROR [4/7] RUN pip install -r requirements.txt                                                                                                                        2.2s
------                                                                                                                                                                          
 > [4/5] RUN pip install -r requirements.txt:
#8 0.727 standard_init_linux.go:219: exec user process caused: exec format error
------
Dockerfile.arm32v7:22
--------------------
  20 |     #     libcurl4-openssl-dev
  21 |     
  22 | >>> RUN pip install -r requirements.txt
  23 |     
  24 |     COPY . .
--------------------
error: failed to solve: rpc error: code = Unknown desc = executor failed running [/bin/sh -c pip install -r requirements.txt]: exit code: 1

My development machine:

• Chip: Intel i7 4510u
• OS: Manjaro KDE

I also tried to build it directly on the Raspberry Pi 4 itself, but alas, it gave me the same errors. I'm at a loss of what to do. Any suggestions?

Answer 1

Oddly enough, it successfully built when I was on a Windows machine, instead of using Manjaro Linux. I still wonder why...