MQA metric service

Computes various metrics for the MQA and stores the results in a database.

Setup

1. Install all of the following software

• Java JDK >= 1.8
• Git >= 2.17
• MongoDB >= 4.0.3

2. Clone the directory and enter it

    git clone [email protected]:european-data-portal/mqa-metric-service.git
   

3. Edit the environment variables in the Dockerfile to your liking. Variables and their purpose are listed below:

Key	Description	Default
PORT	Port this service will run on	8080
MONGODB_SERVER_HOST	MongoDB server address	localhost
MONGODB_SERVER_PORT	MongoDB port	27017
MONGODB_USERNAME	MongoDB user name	null
MONGODB_PASSWORD	MongoDB password
MONGODB_DB_NAME	MongoDB database name	mqa-metrics
PGSQL_SERVER_HOST	PostgreSQL server address, including port and database name	jdbc:postgresql://localhost:5432/mqa
PGSQL_USERNAME	PostgreSQL user name	postgres
PGSQL_PASSWORD	PostgreSQL password	postgres
PGSQL_DEFAULT_LIMIT	Default number of results to return	20
PIVEAU_PIPE_LOG_LEVEL	Log level	INFO

Run

Production

Build the project by using the provided Maven wrapper. This ensures everyone this software is provided to can use the exact same version of the maven build tool. The generated fat-jar can then be found in the target directory.

• Linux

    ./mvnw clean package
    java -jar target/mqa-metric-service-0.1-fat.jar
  

• Windows

    mvnw.cmd clean package
    java -jar target/mqa-metric-service-0.1-fat.jar
  

• Docker

  1. Start your docker daemon
  2. Build the application as described in Windows or Linux
  3. Adjust the port number (EXPOSE in the Dockerfile)
  4. Build the image: docker build -t edp/mqa-metric-service .
  5. Run the image, adjusting the port number as set in step iii: docker run -i -p 8123:8123 edp/mqa-metric-service
  6. Configuration can be changed without rebuilding the image by overriding variables: -e PORT=8124

Development

For use in development two scripts are provided in the project's root folder. These enable hot deployment (dynamic recompiling when changes are made to the source code). Linux users should run the redeploy.sh and Windows users the redeploy.bat file.

Note: The files generated by VertX Codegen may not be detected by your IDE. In this case, mark the directory src/main/generated as Generated Sources Root.

CI

The repository uses the gitlab in-build CI Framework. The .gitlab-ci.yaml file starts as soon a new push event occurs. After running the test cases the application is build, a new docker image is created and stored in the gitlab registry.

API

A formal OpenAPI 3 specification can be found in the src/main/resources/webroot/openapi.yaml file. A visually more appealing version is available at {url}:{port} once the application has been started.

MongoDB schema

All global metrics are stored in single collection. All metrics belonging to a certain catalogue are stored in a separate collection named after the catalogue's ID.