OpenOCR runs on Kubernetes!

• Instructions to run on Google Container Engine

• Instructions to run on Azure Kubernetes Service

If you want to run it on a different Kubernetes provider, particularly on ones that don't offer the Type: LoadBalancer support for Kubernetes Service definitions, you will need to change the open-ocr-httpd service accordingly.

Quick Start - Automated

You can use the bootstrap.sh script found on this directory to automate all of the steps described below.

curl -O bootstrap.sh https://raw.githubusercontent.com/dcasati/open-ocr/dcasati-azure-demo/kubernetes/bootstrap.sh
chmod +x bootstrap.sh

Using the provided script:

usage: bootstrap.sh [-cit]

Options:
    -c     clean up the demo.
    -i     install the demo.
    -t     run a cURL test against the API.

To install the demo:

bootstrap.sh -i

To test the REST API

bootstrap.sh -t

Finally, to clean up all of the resources from the Kubernetes cluster

bootstrap.sh -c

Quick Start - Manual

Create RabbitMQ password

You will want to replace the YOUR_RABBITMQ_PASS below with something more secure.

printf "YOUR_RABBITMQ_PASS" > ./password
kubectl create secret generic rabbit-mq-password --from-file=./password

Clone OpenOCR repo

git clone https://github.com/tleyden/open-ocr.git

Launch RabbitMQ

kubectl create -f kubernetes/pods/rabbitmq.yaml
kubectl create -f kubernetes/services/rabbitmq.yml

Wait until it launches by checking:

kubectl describe pod rabbitmq

and make sure the state is RUNNING

Launch REST API Server

kubectl create -f kubernetes/pods/open_ocr_httpd.yml
kubectl create -f kubernetes/services/open_ocr_httpd.yml

Launch OCR Worker

kubectl create -f kubernetes/replication-controllers/open-ocr-worker.yaml

Verify

Find the external IP

kubectl describe service open-ocr-httpd-service

and look for the LoadBalancer Ingress value. That is the publicly available IP address.

Run curl against the REST API

Replace the IP below with your LoadBalancer Ingress returned above in this command:

curl -X POST -H "Content-Type: application/json" -d '{"img_url":"http://bit.ly/ocrimage","engine":"tesseract"}' http://104.197.33.5/ocr

and you should get the output:

You can create local variables for the pipelines within the template by
prefixing the variable name with a “$" sign. Variable names have to be
composed of alphanumeric characters and the underscore. In the example
below I have used a few variations that work for variable names.