mcltn/food-delivery-satellite-lab

Satellite lab using the food-delivery-app

Satellite Demo - Food Delivery

Food Delivery is a cloud native microservices based application leveraging multiple common cloud native opensource projects to develop a simulated version of an UberEats/GrubHub type business. The application uses OpenShift, Kafka, Redis, and MongoDB as its base set of middleware and adds set of domain specific middleware and an API front end. In this version, as seen in the architecture diagram, the MongoDB instance has been extracted out, and is consumed via a managed service. This lab assumes the MongoDB is already deployed in an IBM Cloud Account.

History and Reference

This set of assets is based on the material used at the Digital Developer Conference: Hybrid Cloud Labs and the IBM Cloud Community Days Conference. You can reference details on the workshops here: https://anthonyamanse.gitbook.io/ibm-satellite-workshop/

The base application itself was developed by Yan Koyfam and Anthony Amaranse and published here: https://github.com/IBM/scaling-apps-with-kafka. It was later modified to remove the dependance on Confluent by Dave Tropeano here: https://github.ibm.com/davetropeano/satellite-demo-food-delivery

This version of the deployment has removed the Mongo Database from running locally in the cluster and is dependent on an existing IBM Cloud Databases for MongoDB instance.

Demo Walkthrough

You can see a walkthrough of the Satellite Config demo utilizing the Food Delivery app here:

Demo Walkthrough

Setup

The main demo script assumes you have a Satellite Location set up with a ROKS cluster.

Create a new project in your cluster (ie: delivery), which will be used as your namespace for the project.

oc new-project <project_name>

The deployments and route-* folders contain the YAML files used to deploy the application. You need to edit and change the {{YOUR_NAMESPACE}} and {{CLUSTER_DOMAIN_NAME}} tokens to use the namespace and cluster URL for your demo system. For this lab, for the routes, we will only use the route-dev/routes.yaml file.

This can be done manually (there are only a few files), you can use Find and Replace to find every instance of {{YOUR_NAMESPACE}} and {{CLUSTER_DOMAIN_NAME}} and replace them in the project directory or use something like mustache to preprocess the YAML files and make the substitutions. If you are using mustache or something similar the file config.yaml is used as a configuration file.

Additionally in the deployments folder, you will need to edit secrets and configmaps. Create a icd-secrets.yaml file from the icd-secrets.yaml.template file and replace the placeholders {{MONGO_HOSTNAME}}, {{MONGO_PORT}}, {{MONGO_USERNAME}} and {{MONGO_PASSWORD}}. Also add your initials or a random character string to the DBNAME variable as well, replacing the {{INITIALS}} placeholder.

Create a icd-cert.yaml file from the icd-cert.yaml.template file and populate the placeholder {{CERTIFICATE}} with the contents from your MongoDB instance.

Apply the configmaps and secrets to your project.

oc create -f icd-cert.yaml

oc create -f icd-secrets.yaml

Create Satellite Cluster Group

• Create a Cloud Satellite cluster group. Add the appropriate cluster to the cluster group.

• Create a Cloud Satellite configuration.

Open the configuration and create a new version using the kafka-redis.yaml file and name it something recognizable such as kafka-redis, repeat this creation of versions for backend.yaml, frontend.yaml, frontend-v2.yaml files in the deployment folder and name the versions appropriately.

• Open the Cloud Satellite configuration for this project and create a version for the routes.yaml file found in the route-dev folder.

• Create a Cloud Satellite subscription for the route labeled with the route's name. choose the associated version and add to the appropriate cluster group(s).

• Create a Cloud Satellite subscription labeled kafka-redis, choose the version kafka-redis and add the cluster group(s).

• Check Pods within the cluster(s) to see if the services are running.

• Create a Cloud Satellite subscription labeled backend-microservices, choose the version backend and add the cluster group(s).

• Check Pods within the cluster(s) to see if the services are running.
• Create a Cloud Satellite subscription labeled frontend, choose the version frontend-v1 or the name you provided for the first frontend yaml file and add the cluster group.
• Check Pods within the cluster(s) to see if the services are running.
• Go to Routes in your cluster(s) and find the example-food route. Open the URL.

• Now we will show updates to the application by switching from a frontend light theme (frontend-v1) to a dark theme (frontend-v2).
• If the frontend-v2 version is already created in your project's configuration, then skip this step. Open the project's Cloud Satellite Configuration and select Versions from the sidenav bar. Find the frontend-v1 version and open the ellipsis to select Duplicate. Change the title to frontend-v2 and the - image: mcltn/example-food-frontend:1.0 to - image: mcltn/example-food-frontend:2.0. Press Add.
• Go into the Cloud Satellite subscriptions. Update the frontend subscription to the frontend-v2 version and press Save.

• Look at Pods to see a new example-food deployment configured (frontend-v2) and another deployment (frontend-v1) terminating.

• Once the frontend-v1 deployment terminates, open the example-food application route again and view the application to see the update to the dark theme that frontend-v2 configured.

• Press Start on the management controls panel in the example-food GUI to increase the Order quantity.
• Adjust the controls to have a higher rate of food orders, increase the speed of the couriers and kitchens.

This demonstration shows the near-to instant update of an application via Cloud Satellite Config.