image-classifier fails with illigal instruction on AMD64 'illegal instruction' when importing tensorflow as tf

[initmahesh]

Please provide us with the following information:

This issue is for a: (mark with an x)

- [ ] bug report -> please search issues before submitting
- [ ] feature request
- [ ] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

Clone this sample Update the .env file with the values for your container registry and make sure that your docker engine has access to it 3.Build the entire solution by opening the control palette (Ctrl+Shift+P), select Build and push IoT Edge Solution (this can take a while...especially to build numpy and pillow...) and select the deployment.test-amd64.template.json manifest file 4.Deploy the solution to your device by right-clicking on the config/deployment.json file, select Create Deployment for Single device and choose your targeted device verify module image-classifier-service are not running. create module image-classifier-service failed in docker.

Any log messages given by the failure

upsquared@UP2:~$ sudo docker ps CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 2153c5a76718 mycapreg.azurecr.io/cameracapture:0.2.8-test-amd64 "python3 -u ./main.py" 6 minutes ago Up 2 seconds 0.0.0.0:5012->5012/tcp camera-capture e8a3815db9c2 mcr.microsoft.com/azureiotedge-hub:1.0 "/bin/sh -c 'echo \"$…" 25 minutes ago Up 4 seconds 443/tcp, 5671/tcp, 8883/tcp edgeHub 0cf73814128b mcr.microsoft.com/azureiotedge-agent:1.0 "/bin/sh -c 'echo \"$…" 7 days ago Up 12 seconds edgeAgent upsquared@UP2:~$ sudo docker ps -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES 2153c5a76718 mycapreg.azurecr.io/cameracapture:0.2.8-test-amd64 "python3 -u ./main.py" 6 minutes ago Up 9 seconds 0.0.0.0:5012->5012/tcp camera-capture e8a3815db9c2 mcr.microsoft.com/azureiotedge-hub:1.0 "/bin/sh -c 'echo \"$…" 25 minutes ago Up 10 seconds 443/tcp, 5671/tcp, 8883/tcp edgeHub e7fae1986de1 mycapreg.azurecr.io/imageclassifierservice:0.2.5-amd64 "/bin/sh -c 'python …" 25 minutes ago Exited (132) 4 minutes ago image-classifier-service 0cf73814128b mcr.microsoft.com/azureiotedge-agent:1.0 "/bin/sh -c 'echo \"$…" 7 days ago Up 19 seconds edgeAgent upsquared@UP2:~$ sudo docker logs -f image-classifier-service --tail 200 Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) Illegal instruction (core dumped) upsquared@UP2:~$ sudo systemctl stop iotedge

upsquared@UP2:~$ sudo docker run -it --rm mycapreg.azurecr.io/imageclassifierservice:0.2.5-amd64 Illegal instruction (core dumped) upsquared@UP2:~$ sudo docker run -it --rm mycapreg.azurecr.io/imageclassifierservice:0.2.5-amd64 bash

---

_ /____ __/ /____ / _ _ _/ _ _/ / _ / _ | /| / / / / / / / /( )/ // / / _ / / / // / |/ |/ / // _/// //// ___/// // // \/__/|/

WARNING: You are running this container as root, which can cause new files in mounted volumes to be created as the root user on your host machine.

To avoid this, run the container by specifying your user's userid:

$ docker run -u $(id -u):$(id -g) args...

root@4a0538a0028b:/app# python3 Python 3.6.8 (default, Aug 20 2019, 17:12:48) [GCC 8.3.0] on linux

import tensorflow as tf Illegal instruction (core dumped)

Expected/desired behavior

OS and Version?

Linux ubuntu 18.04

Versions

Mention any other details that might be useful

---

Thanks! We'll be in touch soon.

[initmahesh]

The issue seems to be that there is no AVX support Intel UP2 AI Edge Kit as per here (https://stackoverflow.com/questions/49094597/illegal-instruction-core-dumped-after-running-import-tensorflow). AVX is needed by TF beyond build 1.6 , if we go to build 1.5 we can get pass this issue but then operation tensorflow.compat.v1 is not supported as that got added in 1.14 onwards ….

[emmanuel-bv]

Fixed with https://github.com/Azure-Samples/Custom-vision-service-iot-edge-raspberry-pi/pull/62