C4-W1-Bind Mount Error - Ungraded lab: First look at Tensorflow Serving with Docker

Course Q&A Machine Learning Engineering for Production(MLOps) Deploying Machine Learning Models in Production
1

Hello,

I have been trying many things but can’t get this issue resolved. For the ungraded lab “First look at Tensorflow Serving with Docker”, I keep getting the error message below in my wsl shell. You can also see from the attached picture that the files and path are valid, despite what the error message shows (please see the final lines in the picture). Please let me know if anyone could resolve this. Thanks.

Regards,
Matt

"docker: Error response from daemon: invalid mount config for type "bind": bind source path does not exist: /tmp/tfserving/serving/tensorflow_serving/servables/tensorflow/testdata/saved_model_half_plus_two_cpu.
See 'docker run --help'.

[1]+  Done(125)                  docker.exe run --rm -p 8501:8501 --mount type=bind,source=/tmp/tfserving/serving/tensorflow_serving/servables/tensorflow/testdata/saved_model_half_plus_two_cpu,target=/models/half_plus_two -e MODEL_NAME=half_plus_two -t tensorflow/serving"

1
11473×373 19.3 KB

3

Please share your outputs in text format.

4

I did. It is right above the picture in the quotation marks.

5

I was referring to the entire output. Sharing the terminal output as text instead of screenshot helps verification easier.
The lab runs fine on linux. I’ve asked someone with a windows machine to reply to this topic on the mentor forum. Please upvote this topic.

6

Hi Matt! Sorry my Windows machine is not properly set up at the moment but I think you will have to modify the source path to match the expected format in that OS. Please see the Windows related troubleshooting topics here, particularly the one that talks about path conversion. I did a similar fix before when I was setting up Kubernetes on Windows. I’ll try to come back to this thread if I can get my old Windows laptop back. Hope this helps!

7

Please see below. This time, I put it in another location but still got the same error message:

mahadi@moham:/var$ docker.exe run --rm -p 8501:8501 --mount type=bind,source=/var/tmp/saved_model_half_plus_two_cpu,target=/models/half_plus_two -e MODEL_NAME=half_plus_two -t tensorflow/serving &
[1] 236
mahadi@moham:/var$ docker: Error response from daemon: invalid mount config for type “bind”: bind source path does not exist: /var/tmp/saved_model_half_plus_two_cpu.
See ‘docker run --help’.

[1]+ Exit 125 docker.exe run --rm -p 8501:8501 --mount type=bind,source=/var/tmp/saved_model_half_plus_two_cpu,target=/models/half_plus_two -e MODEL_NAME=half_plus_two -t tensorflow/serving
mahadi@moham:/var$ cd tmp/saved_model_half_plus_two_cpu
mahadi@moham:/var/tmp/saved_model_half_plus_two_cpu$ pwd
/var/tmp/saved_model_half_plus_two_cpu
mahadi@moham:/var/tmp/saved_model_half_plus_two_cpu$

8

Hi Chris,

Thanks for pointing me to the right direction. Path was one of the issues. The other issue was the changes and some quick fixes I had made myself to the Linux kernel to make it work for the labs (changed the root and some other things which led to some super-weird behavior). After I re-installed everything and used the right path, everything worked out. Thanks again!

Regards,
Matt

1 Like
9

Awesome! Glad it worked!

10

I could resolve the error by copying the “tfserving” directory from /tmp to my home directory in ubuntu; then I ran the following command and it worked:

docker run --rm -p 8501:8501 --mount type=bind,source=/home/shayan/tfserving/serving/tensorflow_serving/servables/tensorflow/testdata/saved_model_half_plus_two_cpu,target=/models/half_plus_two -e MODEL_NAME=half_plus_two -t tensorflow/serving