AWS Batch Container Error: "ResourceInitializationError no users found"

Question

When triggering a batch job (Fargate Job Queue), the status is going to **FAILED** with the following error message:

> Cannotstartcontainererror: ResourceInitializationError: unable to create new container: mount callback failed on /tmp/containerd-mount3975084381: no users found

Unfortunately I can't find any similar errors online.
For reference, the Dockerfile that I'm building is simply the following:
```Docker
FROM python:3.8-slim-buster
WORKDIR /app
USER root

COPY requirements.txt requirements.txt
RUN pip3 install -r requirements.txt

COPY . .

CMD [ "python3", "run.py"]
```
And the contents of run.py are as follows:
```python
print("Python script has run!")
```

The only other file in the image is requirements.txt, which contains just the line `requests`.

I temporarily gave the ecs runtime role full admin permissions to see if that would help, but the error still occurred (so I changed it back).

Any help here would be appreciated!

Accepted Answer

Found my answer! The job definition had the user set to `ubuntu`, which wasn't available in the `python:3.8-slim-buster` image. I changed it to `root` and it works perfectly.

Answer

Not enough info to help with this. I would try these things:
1. I do not think it is an issue with your docker image, but anyway, it is better to make sure it works as expected. Try to build a docker image locally and run this `docker run YOUR_IMAGE python3 run.py`
2. check CloudWatch logs. There must be a stream with the title `Batch...`, something like that. Maybe there will be more info about the error
3. try to use not Fargate, but EC2. If it work, the issue probably with task configuration

Feel free to ping me if you need more help, I am curious why this thing does not work :)

AWS Batch Container Error: "ResourceInitializationError no users found"

相关内容