Deploying R Machine Learning Model with Plumber and Azure Kubernetes Service

Deploying and Serving R Models to AKS with Plumber

Model code adapted from: https://shirinsplayground.netlify.com/2018/01/plumber/
Author of model: Dr. Shirin Glander

Azure Kubernetes Service (AKS) can be used to deploy and serve Machine Learning Models. R is a popular language for Machine Learning and using Plumber (https://www.rplumber.io/) you can construct an REST API to expose the Machine Learning model. Using Docker, the R model and the Plumber script can be containerized using Docker and served on Azure Kubernetes Service. The image is deployed through a Deployment object and exposed to the Internet thourgh a LoadBalancer object.

How to deploy this example

Pre-requisites

• Azure Subscription
• Azure Kubernetes Service Cluster
• Azure Container Registry
• Azure CLI
• kubectl and helm tools installed
• Docker

Instructions

1. Clone this repository onto your local machine

git clone https://github.com/mpfishe2/r-deploy-model-aks.git

2. Make the cloned repository your working directory

cd r-deploy-model-aks`

3. Pull the Docker image mpfplumber

docker pull mpfishe2/mpfplumber

4. Tag the image to prepare it for pushing to your Azure Container Registry

docker tag mpfishe2/mpfplumber:latest <name-of-acr-registry>.azurecr.io/mpfplumber:latest

Note: If the name of the Azure Container Registry is containerland than the tag would be containerland.azurecr.io/mpfplumber:latest

5. Login in to Azure CLI

az login

6. Login in to your Azure Container Registry

az acr login --name <name-of-acr-registry>

7. Push the image that you tagged in Step 4 to your Azure Container Registry

docker push <name-of-acr-registry>.azurecr.io/mpfplumber:latest

8. Deploy the application in Azure Kubernetes Service Cluster

Helm Deployment

helm install ./raksdeploy

Non-Helm Deployment

cd aksfiles
kubectl apply -f .

9. Get the <EXTERNAL-IP> of the LoadBalancer that is Deployed (Note: It might take a few minutes for the External IP to populate)

kubectl get svc

The above command will display all the services. You are looking for the one associated with your Helm installation or one that is titled aksdemo-model-deploy

10. Test using cURL

curl "http://<EXTERNAL-IP-FROM-SERVICE>/predict?age=0.511111111111111&bp=0.111111111111111&sg_1.005=1&sg_1.010=0&sg_1.015=0&sg_1.020=0&sg_1.025=0&al_0=0&al_1=0&al_2=0&al_3=0&al_4=1&al_5=0&su_0=1&su_1=0&su_2=0&su_3=0&su_4=0&su_5=0&rbc_normal=1&rbc_abnormal=0&pc_normal=0&pc_abnormal=1&pcc_present=1&pcc_notpresent=0&ba_present=0&ba_notpresent=1&bgr=0.193877551020408&bu=0.139386189258312&sc=0.0447368421052632&sod=0.653374233128834&pot=0&hemo=0.455056179775281&pcv=0.425925925925926&wbcc=0.170454545454545&rbcc=0.225&htn_yes=1&htn_no=0&dm_yes=0&dm_no=1&cad_yes=0&cad_no=1&appet_good=0&appet_poor=1&pe_yes=1&pe_no=0&ane_yes=1&ane_no=0"

This should be the response:

----------------
Test case predicted to be ckd
----------------