Infinispan SQL Store Demo with Spring-Boot and Quarkus

Description

This demo illustrates how to build a high-performance product catalogue by integrating a traditional SQL-based backend with a distributed in-memory cache using Infinispan.

• Back-office REST API, traditional SQL database, manages product catalogues.
• Front-office REST API use Infinispan SQL Cache Stores to load and serve product catalogue and sales data directly from memory.

Retail Catalogue

Retail Catalogue is a very simple Quarkus application that persists data in a Postgresql database. In dev mode, runs under localhost:8080

'commands' endpoint displays the content of the database model.

[
    {
        "id": 5,
        "buyer": {
            "id": 4,
            "country": "Spain",
            "email": "[email protected]",
            "name": "Maria Rosario Frechilla",
            "phone": "+34 666"
        },
    "products": [
            {
                "id": 1,
                "code": "c123",
                "name": "Skirt Party",
                "price": 50,
                "stock": 20
            },
            {
                "id": 2,
                "code": "c456",
                "name": "Pants Party",
                "price": 20,
                "stock": 10
            },
            {
                "id": 3,
                "code": "c789",
                "name": "Dress Party",
                "price": 90,
                "stock": 20
            }
            ],
          "promotion": true
        }
]

Inmemory Catalogue

There are two implementations of the service:

• Quarkus version : inmemory-catalogue-quarkus
• Spring Boot 3 version: inmemory-catalogue-spring-boot

Overall Description

Inmemory Catalogue creates two SQL Cache Stores with Infinispan on startup.

1. catalogue-table-store is a table sql store persisted cache that maps to the RetailProduct table
2. sold-products-query-storeis a query sql store persisted cache that joins information across multiple tables.

Both caches use indexing.

The application exposes 3 endpoints:

• catalogue: Lists the catalogue content. Can filter by name, stock units and prices using query parameters.
• catalogue/{code}: Displays the catalogue product by code
• sales: Lists the products than have been sold. Filter can be done name of the product and country.

Dev mode

• Run Infinispan and Postgresql in the same network with docker

docker-compose up

Quarkus

Dev services are disabled since the Infinispan and Postresql need to be in the same docker network and Infinispan access the Postgresql database directly to load the data into the caches.

The application runs in localhost:8180

• Run Retail Catalogue and Store

cd infinispan-sqlstore-demo/retail-catalogue 
./mvnw quarkus:dev 

• Run Infinispan In-memory catalogue

cd infinispan-sqlstore-demo/inmemory-catalogue-quarkus 
./mvnw quarkus:dev 

Spring Boot

The application runs in localhost:8180.

• Run Retail Catalogue and Store

cd infinispan-sqlstore-demo/retail-catalogue 
./mvnw quarkus:dev 

• Run Infinispan In-memory catalogue

cd infinispan-sqlstore-demo/inmemory-catalogue-spring-boot
./mvnw spring-boot:run 

Links

Additional resources and useful links:

• Infinispan Query and Indexing Guide
• Infinispan SQL Cache Storage Documentation
• Quarkus Infinispan Extension

Deploy Minikube

1. Install and start Minikube

 minikube start --driver=virtualbox --cpus 4 --memory "8192mb"

2. Create Kubernetes Secret sqlstore-credentials

kubectl create secret generic sqlstore-credentials --from-literal=db-username=infinispan --from-literal=db-password=secret --from-literal=infinispan-username=admin --from-literal=infinispan-password=secret

3. Deploy Postgresql

kubectl apply -f deploy/postgres.yaml

4. Build and deploy Retail Catalogue

eval $(minikube -p minikube docker-env)
cd retail-catalogue
 .mvnw clean package -Dquarkus.kubernetes.deploy=true -DskipTests=true 

# Export the service URL 
export RETAIL_CATALOGUE=$(minikube service --url retail-store-service) 

# Check the commands endpoint 
curl $RETAIL_CATALOGUE/commands

5. Install the Infinispan Operator

curl -sL https://github.com/operator-framework/operator-lifecycle-manager/releases/download/v0.31.0/install.sh | bash -s v0.31.0
kubectl create -f https://operatorhub.io/install/infinispan.yaml
kubectl get csv -n operators

You should see something like

NAME                          DISPLAY               VERSION   REPLACES                      PHASE
infinispan-operator.v2.4.12   Infinispan Operator   2.4.12    infinispan-operator.v2.4.11   Succeeded

6. Create Infinispan identities secret

kubectl create secret generic --from-file=identities.yaml connect-secret

7. Deploy infinispan server

kubectl apply -f infinispan.yaml

You should see something like the following list

 minikube service list    
                                                                                    15:17:03
|-------------|-----------------------|--------------|-----------------------------|
|  NAMESPACE  |         NAME          | TARGET PORT  |             URL             |
|-------------|-----------------------|--------------|-----------------------------|
| default     | infinispan            | No node port |
| default     | infinispan-admin      | No node port |
| default     | infinispan-external   |        11222 | http://192.168.59.101:30000 |
| default     | infinispan-ping       | No node port |
| default     | kubernetes            | No node port |
| default     | postgres              | http/5432    | http://192.168.59.101:31684 |
| default     | retail-store-service  | http/80      | http://192.168.59.101:31056 |
| kube-system | kube-dns              | No node port |
| olm         | operatorhubio-catalog | No node port |
| olm         | packageserver-service | No node port |
|-------------|-----------------------|--------------|-----------------------------|

8. Build and deploy Inmemory Catalogue Quarkus

eval $(minikube -p minikube docker-env)
cd inmemory-catalogue-quarkus
./mvnw clean package -Dquarkus.kubernetes.deploy=true -DskipTests=true 

# Export the service URL 
export INMEMORY_CATALOGUE=$(minikube service --url inmemory-catalogue-service) 

curl $INMEMORY_CATALOGUE