Customized R container

[aliraza0337]

1: I downloaded the clipper, git clonned. 2: I added some extra lines to this file in order to customize the r-container 3: I re-install the clipper_admin sudo pip install clipper_admin 4: I re-install the R package (Rclipper) 5: I run my R model 6: I run my clipper_admin (python code) 7: I do docker logs --tail 100 -f Hoping to see my libraries being downloaded and installed. BUT I don't! Any help will be greatly appreciated.

[csmithnm]

I believe you would need to also change this line to point to your new R image.

#If your new container name is "clipper/r-container-base:custom" here is what the line would look like: CLIPPER_R_CONTAINER_BASE_IMAGE = "clipper/r-container-base:custom")

OPTION 2: In trying to stay more with the Rclipper code, I renamed the Rclipper Docker image clipper/r-container-base:develop to clipper/r-container-base:original. Then named the new Docker image, that I built using an altered version of the [RContainerDockerfile] (https://github.com/ucbrise/clipper/blob/develop/dockerfiles/RContainerDockerfile), to clipper/r-container-base:develop. And finally did the ipython deploy_model. But when I did that final step the deploy got hung up. So I hope you have better luck.

#Change the image name of the Rclipper original docker image to save it off docker tag clipper/r-container-base:develop clipper/r-container-base:original

#Change the image name of your new docker image to the one the Rclipper recognizes. docker tag <your-docker-image> clipper/r-container-base:develop