order-service

Microservice that is a part for an Ingestor demonstration. The user using the acme-ui micro-front creates a datasource for new orders. The python ingestor microservice reads a cvs file parse it using Spark, validates the fields (according to the rules defined in the datasource) and the correct record send a kafka event that it is listened by this microservice.

The order-service depends on Kafka events generated by the Python ingestor microservice.

Introduction

This is a microservice for demo purpose to test de Ingestor microservice in a high volume fashion transactions.

This microservice receives a bunch of Kafka events from PhySpark to insert new orders in a Neo4j database (in this case is the acmeDB);. We test how this microservice behaves with volume of transactions.

And also has the functionality to query the orders in the acme-ui microservice.

note: The purpose is to create a table for time response for thousands of transaction with more than one spark-workers instances.

Send Events

No events are sent via Kafka it is just a listener for ordenes topic.

Databases

No database is used directly for this microservice:

The bupNeo4j is handled by the bup-service microservice.

Orders queries from Neo4jDB

This microservice utilizes bupDB to store in the BUP database, GraphQL as an API to query parameters a Spring GraphQL and Neo4j GraphQL Java library.

note: NOT use Spring Data and Spring Neo4j Data libraries (i.e., repositories). All is done in GraphQL which we believe is much better solution.

For more information see example:

graphql-spring-boot or visit link:

https://github.com/neo4j-graphql/neo4j-graphql-java/tree/master/examples/graphql-spring-boot

Or the official repository (see examples) for Spring for GraphQL

https://github.com/spring-projects/spring-graphql

Orders table

For simplicity of the example and not create a new RDBMS we use Neo4j to store the orders table. No relationships exist for this table.

Schema

For developers, we can generate at compile time the schema augmentation for debug purpose. This is done with the neo4j-graphql-augmented-schema-generator-maven-plugin. Since there is not Gradle plugin you can do the following process:

(a) In Gradle use the maven-publish plugin with the publish task goal. When you run:

./gradlew publish

A pom file is created in the directory ./build/publications with the name pom-default.xml.

(b) The file was copied in the project directory and modified to include the neo4j-graphql-augmented-schema-generator-maven-plugin and comment some dependencies from the maven local repository (i.e., ailegorreta-kit)

mvn -f pom-default.xml compile

To create tha augmented schema in the directory /target/augmented-schem/neo4j.graphql.

note: This process has to be done every time the schema is modified.

Events listener

Listener any event with the topic ordenes

Running on Docker Desktop

Create the image manually

./gradlew bootBuildImage

Publish the image to GitHub manually

./gradlew bootBuildImage \
   --imageName ghcr.io/rlegorreta/order-service \
   --publishImage \
   -PregistryUrl=ghcr.io \
   -PregistryUsername=rlegorreta \
   -PregistryToken=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

Publish the image to GitHub from the IntelliJ

To publish the image to GitHub from the IDE IntelliJ a file inside the directory .github/workflows/commit-stage.yml was created.

To validate the manifest file for kubernetes run the following command:

kubeval --strict -d k8s

This file compiles de project, test it (for this project is disabled for some bug), test vulnerabilities running skype, commits the code, sends a report of vulnerabilities, creates the image and lastly push the container image.

For detail information see .github/workflows/commit-stage.yml file.

Run the image inside the Docker desktop

docker run \
    --net ailegorretaNet \
    -p 8530:8530 \
    -e SPRING_PROFILES_ACTIVE=local \
    order-service

Or a better method use the docker-compose tool. Go to the directory ailegorreta-deployment/docker-platform and run the command:

docker-compose up

Run inside Kubernetes

Manually

If we do not use the Tilttool nd want to do it manually, first we need to create the image:

Fist step:

./gradlew bootBuildImage

Second step:

Then we have to load the image inside the minikube executing the command:

image load ailegorreta/order-service --profile ailegorreta

To verify that the image has been loaded we can execute the command that lists all minikube images:

kubectl get pods --all-namespaces -o jsonpath="{..image}" | tr -s '[[:space:]]' '\n' | sort | uniq -c\n

Third step:

Then execute the deployment defined in the file k8s/deployment.yml with the command:

kubectl apply -f k8s/deployment.yml

And after the deployment can be deleted executing:

kubectl apply -f k8s/deployment.yml

Fourth step:

For service discovery we need to create a service applying with the file: k8s/service.yml executing the command:

kubectl apply -f k8s/service.yml

And after the process we can delete the service executing:

kubectl deltete -f k8s/service.yml

Fifth step:

If we want to use the project outside kubernetes we have to forward the port as follows:

kubectl port-forward service/order-service 8530:80

Appendix:

If we want to see the logs for this pod we can execute the following command:

kubectl logs deployment/order-service

Using Tilt tool

To avoid all these boilerplate steps is much better and faster to use the Tilt tool as follows: first create see the file located in the root directory of the project called TiltFile. This file has the content:

# Tilt file for order-service
# Build
custom_build(
    # Name of the container image
    ref = 'order-service',
    # Command to build the container image
    command = './gradlew bootBuildImage --imageName $EXPECTED_REF',
    # Files to watch that trigger a new build
    deps = ['build.gradle', 'src']
)

# Deploy
k8s_yaml(['k8s/deployment.yml', 'k8s/service.yml'])

# Manage
k8s_resource('order-service', port_forwards=['8530'])

To execute all five steps manually we just need to execute the command:

tilt up

In order to see the log of the deployment process please visit the following URL:

http://localhost:10350

Or execute outside Tilt the command:

kubectl logs deployment/order-service

In order to undeploy everything just execute the command:

tilt down

To run inside a docker desktop the microservice need to use http://order-service:8530 to 8530 path

Reference Documentation

Links to Springboot 3 Observability

https://tanzu.vmware.com/developer/guides/observability-reactive-spring-boot-3/

Baeldung:

https://www.baeldung.com/spring-boot-3-observability

Contact AI Legorreta

Feel free to reach out to AI Legorreta on web page.

rlegorreta / order-service Goto Github PK

order-service's Introduction