Apigee Edge Microgateway on Docker & Kubernetes

This project describe how you can setup multiple Apigee Edge Microgateways using proxy filters, docker and Kubernetes.

Prerequisites

  1. Docker installation. For more info please visit Docker home page
  2. Account with Apigee and understanding how to setup Edge Microgateway proxies. Please see Edge Microgateway Documentation for more information
  3. Node.js 4.x or later. For more info visit Nodejs Github Release Document
  4. Basic understanding and experience of Docker
  5. If you choose to run it in a Kubernetes environment, you will need to have some experience with Kubernetes

Section 1 - Configuration Preperation

This section is to set up two Edge Microgateway aware proxies to demonstrate how we can deploy seperate microgateway containers to be used for seperate proxy endpoints by utilizing proxy filters.

  1. Create two Microgateway aware reverse proxies in your existing Apigee portal:
    • edgemicro_firstproxy
      • Proxy Name: edgemicro_firstproxy
      • Proxy Base Path: firstproxy
      • Existing API http://httpbin.org/
    • edgemicro_secondproxy
      • Proxy Name: edgemicro_secondproxy
      • Proxy Base Path: secondproxy
      • Existing API http://httpbin.org/

  2. In this step, you will generate a key, secret and the {org}-{env}-config.yaml for later use. This section assumes you already have node.js and npm installed on your localhost or server.

    Please open a terminal in your localhost/server and do the following:

    • Install Edge microgateway
     npm install -g edgemicro
    • Initialize Edge microgateway
     edgemicro init
    • Configure Edge microgateway and save the key and secret for later use.
     edgemicro configure -o "your-orgname" -e "your-envname" -u "your-username"

    • If the configuration is successful, you will see a key, secret and token being displayed. The {org}-{env}-config.yaml file has also been generated in the ~/.edgemicro directory.

    • Save the key, secret and token in a text file of your choice for later use.

Section 2 - Build docker images

The steps in this section are to build Docker container images. To complete this section, you will need a Docker hub account(or equivalent).

  1. In your localhost or server, change directory to .edgemicro. Run cd ~/.edgemicro/

  2. Clone the project git clone https://github.com/zcai2672/apigee-edgemicro-docker.git

  3. Copy the {org}-{env}-config.yaml generated in the Section 1 - Configuration Preperation to the ‘apigee-edgemicro-docker’ folder created from the previous step.

  4. Switch directory cd apigee-edgemicro-docker

  5. Add proxyPattern: edgemicro_firstproxy* to the edge_config member in the {org}-{env}-config.yaml 
     edge_config:
   proxyPattern: edgemicro_firstproxy*
   .
   .
   . omitted for brevity

  6. Build the docker image using following command: docker build --build-arg ORG="your-orgname" --build-arg ENV="your-env" -t microgateway:first .

  7. Run docker images to get the IMAGE_ID of the container image created in previous step.

  8. Tag the image docker tag <image_id> <docker_hub_id>/microgateway:first

  9. Login to docker hub, run docker login

  10. Push the image to a container register such as Docker Hub (you can also use Google Container Register) docker push <docker_hub_id>/microgateway:first

  11. Now edit the {org}-{env}-config.yaml file again by changing the proxyPattern configuration to ‘edgemicro_secondproxy*’ proxyPattern: edgemicro_secondproxy* in the edge_config member in the {org}-{env}-config.yaml 
    edge_config:
  proxyPattern: edgemicro_secondproxy*
  .
  .
  . omitted for brevity

  12. Build the docker image using following command: docker build --build-arg ORG="your-orgname" --build-arg ENV="your-env" -t microgateway:second .

  13. Run docker images to get the IMAGE_ID of image created in previous step.

  14. Run docker tag <image_id> <docker_hub_id>/microgateway:second

  15. push the images to a container register such as Docker Hub or Google Container register docker push <docker_hub_id>/microgateway:second

Section 3 - Deployments

You have two deployment options in this section, You can choose to deploy docker only in option 1 or you can deploy your images in a Kubernetes environment.

Option 1 - Docker only deployment

For this option, you will be deploying two Docker container images, microgateway:first in port 8000 and microgateway:second in port 8001

You can run the the microgateway image in same local environment that you have previously set up.

  1. To start running microgateway:first image 
    docker run -d -p 8000:8000 -e EDGEMICRO_ORG="your-orgname" -e EDGEMICRO_ENV="your-env" -e EDGEMICRO_KEY="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" -e  EDGEMICRO_SECRET="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" -P -it microgateway:first

  2. To test, run curl http://localhost:8000/firstproxy/anything

    If deployment is successful, you should see something like this

     {"error":"missing_authorization","error_description":"Missing Authorization header"}

    To test if the filter is working, run curl http://localhost:8000/secondproxy/anything You should get the following if the proxy filter that you have configured works.

     {"message":"no match found for /secondproxy/anything","status":404}

  3. Deploy the second gateway microgateway:second image docker run -d -p 8001:8000 -e EDGEMICRO_ORG="your-orgname" -e EDGEMICRO_ENV="your-env" -e EDGEMICRO_KEY="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" -e EDGEMICRO_SECRET="xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx" -P -it microgateway:second

  4. To test, run curl http://localhost:8001/secondproxy/anything

    you should see something like this

     {"error":"missing_authorization","error_description":"Missing Authorization header"}

    Then run curl http://localhost:8001/firstproxy/anything to test the filter you have configured for this proxy. You should get the following if your filter works. {"message":"no match found for /firstproxy/anything","status":404}

Option 2 - Docker in GCP Kubernetes deployment

You will need to a Google Cloud platform Account for this option. A local Kubernetes installation should also work but has not been tested at this stage.

  1. Login to your GCP console, create your Kubernetes cluster under Kubernetes Engine section.

  2. Run the following command to clone the Microgate way project git clone https://github.com/zcai2672/apigee-edgemicro-docker.git

  3. Switch directory cd apigee-edgemicro-docker

  4. Create Kubernetes secret
    • Use the key and secrets generated in step 2 of Section 1 - Configuration Preperation and it to to base64 Convert each of these values into base64. THis will help store those credentails into Kubernetes secrets. 
      echo -n "org" | base64
echo -n "env" | base64
echo -n "key" | base64
echo -n "secret" | base64

    • Use the results in the configuration in the mgw-secret.yaml

      apiVersion: v1
kind: Secret
metadata:
  name: mgwsecret
type: Opaque
data:
  mgorg: xxxxxx
  mgenv: xxxxxx
  mgkey: xxxxxx
  mgsecret: xxxxxx

      Run this file in your kubernetes environment to create the secret configuration

       kubectl create -f  mgw-secret.yaml

      For more info on Kubernetes secrets, please see Kubernetes Secret Doc

      You can check if the secret creation by issuring the following command

      kubectl get secrets

      You will see something similar to this:

      NAME           TYPE       DATA      AGE
mgwsecret      Opaque     4         23h

  5. Create Kubernetes microgateway pods.

    Open mgw-pod-first.yaml and mgw-pod-second.yaml. Update the placeholder for the ‘image’ member to point to the correct docker container registry location (In Step 8) for file mgw-pod-first.yaml and mgw-pod-second.yaml files.

    If you are using Docker hub, make sure you are logged in

    docker login

    Run the script to create the pods

    kubectl create -f mgw-pod-first.yaml

    kubectl create -f mgw-pod-second.yaml

    Check if the pods has been created

    kubectl get pods

    You will see something like this:

    NAME                                  READY     STATUS    RESTARTS   AGE
edge-microgateway-first               1/2       Running   0          17h
edge-microgateway-second              1/2       Running   0          17h

  6. Create Kubernetes microgateway Service to expose the gateway

    Make sure that your selector is configured to point to the pod labels member from the step above.

    apiVersion: v1
kind: Service
metadata:
  name: edge-microgateway
  labels:
    app: edge-microgateway
spec:
  ports:
  - port: 8000
    name: http
  selector:
    app: edge-microgateway

    Check if service has been created

    kubectl get services

    You should see something like this:

    NAME                TYPE           CLUSTER-IP      EXTERNAL-IP     PORT(S)          AGE
edge-microgateway   LoadBalancer   xx.xx.xx.xx     xx.xx.xx.xx     8000:31827/TCP   17h
kubernetes          ClusterIP      10.59.240.1     <none>          443/TCP          1d

    To test the service endpoint, please make sure you are using a terminal on a seperate server or your local machine. Use generated external ip obtained from the previous step and run:

    curl -v http://<EXTERNAL-IP>:8000/firstproxy/anything -v

    If the deployment is successful you should be see something like this:

    {"error":"missing_authorization","error_description":"Missing Authorization header"}