Skip to content

ideas-into-software/automated-linguistic-analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A.L.A. (Automated Linguistic Analysis) application

Sample application showcasing usage of technologies such as:

  • OSGi R7 Promises for asynchronous generation of transcriptions and linguistic analyses
  • OSGi R7 Push Stream and JAX RS Server Sent Events for push notifications of processing status
  • Apache Camel 2.23.1 and RabbitMQ 3.7 for asynchronous communication between services
  • JPA 2.1 and Hibernate 5.2.12, along with OSGi R7 JPA and Transaction Control services, for persistence layer
  • OSGi R7 HTTP and JAX RS Whiteboard for registering servlets, resources and REST controllers
  • OSGi R7 Configurator, Configuration Admin and Metatype services for automatic configuration of components
  • OSGi R7 Declarative Services for dependency injection
  • Maven automated build of Docker images
  • Maven automated deployment into Kubernetes cluster
  • RabbitMQ message broker as a StatefulSet
  • CockroachDB relational database as a StatefulSet

For detailed overview, please see the following articles:

Configuring

I. Mono version

1. Set up PostgreSQL database

  • Create database and user using PostgreSQL command line or GUI

  • If not already installed, install the postgresql-contrib package as that’s what’s required to use the gen_random_uuid() function; same function is available out-of-the-box on CockroachDB which is used for Kubernetes version

2. Set up RabbitMQ message broker

  • (optional) Set up RabbitMQ Web Management console – once set up it will be accessible via http://localhost:15672/ (u/p: guest/guest)

    rabbitmq-plugins enable rabbitmq_management

3. Open configuration.json file residing in mono-app module in src/main/resources/OSGI-INF/configurator directory

  • Under org.apache.aries.tx.control.jpa.xa~ala PID, configure database, user name and password
  • (optional) Under software.into.ala.service.messaging PID, configure message broker host and port or leave as is to use defaults
  • Under org.apache.aries.jax.rs.whiteboard.default PID, configure file storage location – this must match file storage location seen in software.into.ala.service.transcription PID
  • Under software.into.ala.service.transcription PID, configure IBM Watson Speech to Text endpoint and API key – small footprint instances are free and sufficient to test drive this application; for more information see https://www.ibm.com/watson/services/speech-to-text/
  • Under software.into.ala.service.linguistics PID, configure IBM Watson Tone Analyzer endpoint and API key – small footprint instances are free and sufficient to test drive this application; for more information see https://www.ibm.com/watson/services/tone-analyzer/

II. Kubernetes version

1. Web application (k8-web-app module) – no configuration is necessary, unless you wish to use custom settings, i.e.

  • Open configuration.json file residing in k8-web-app module in src/main/resources/OSGI-INF/configurator directory
  • Unless you wish to use different settings, DO NOT modify configuration for org.apache.aries.tx.control.jpa.xa~ala PID – in Kubernetes version, both user and database are created automatically during cluster initialization, which also happens automatically during deployment
  • Unless you wish to use different settings, DO NOT modify configuration for software.into.ala.service.messaging PID
  • Unless you wish to use different settings, DO NOT modify configuration for org.apache.aries.jax.rs.whiteboard.default PID, e.g. file storage location

2. Transcription application (k8-transcription-app module) – only IBM Watson configuration is required, i.e.

  • Open configuration.json file residing in k8-transcription-app module in src/main/resources/OSGI-INF/configurator directory
  • Under software.into.ala.service.transcription PID, configure IBM Watson Speech to Text endpoint and API key – small footprint instances are free and sufficient to test drive this application; for more information see https://www.ibm.com/watson/services/speech-to-text/

3. Linguistics application (k8-linguistics-app module) – only IBM Watson configuration is required, i.e.

  • Open configuration.json file residing in k8-linguistics-app module in src/main/resources/OSGI-INF/configurator directory
  • Under software.into.ala.service.linguistics PID, configure IBM Watson Tone Analyzer endpoint and API key – small footprint instances are free and sufficient to test drive this application; for more information see https://www.ibm.com/watson/services/tone-analyzer/

4. In main project POM

  • Modify the docker-registry property to use external Docker registry or set up a local Docker registry; for more information see https://docs.docker.com/registry/deploying/
  • If not using local Docker registry, remove the k8-maven.insecureRegistry property – this one is used for minikube automation and required if using local Docker registry
  • (optional) Adjust the k8-maven.memory and k8-maven.cpus properties as needed – this is used for minikube automation

5. Set up and start Kubernetes cluster – e.g. minikube can be used to quickly set up local Kubernetes cluster; for more information see https://github.com/kubernetes/minikube

6. (optional) Customize Kubernetes descriptors for deploying cluster versions of RabbitMQ and CockroachDB as StatefulSets – these reside in k8-infra project, in src/main/k8 directory; for more information see:

Building

I. Mono version

mvn install

II. Kubernetes version

Build project and Docker images, automatically pushing Docker images to configured repository

mvn -P k8 install

Deploying

I. Mono version

1. Go to mono-app module directory and run

java -jar target/mono-app.jar

2. Open http://localhost:8080/spa/index.html in web browser

II. Kubernetes version

1. To deploy onto Kubernetes cluster from main project directory, run

mvn -P k8 deploy

2. Find IP and service port of ala-k8-web-app service – e.g. for minikube you can use

minikube service --url ala-k8-web-app

3. Open http://IP:SERVICEPORT/spa/index.html in web browser

Architecture

I. Mono version

1. Deployment view

Mono Deployment view

2. Component view

Mono Component view

II. Kubernetes version

1. Deployment view

Kubernetes Deployment view

2. Component view

  • Web application

Kubernetes Web application Component view

  • Transcription application

Kubernetes Web application Component view

  • Linguistics application

Kubernetes Web application Component view

Technology stack

  • Java 8

  • OSGi Release 7 Core and Enterprise specifications implementations, including

    • OSGi Configurator
    • OSGi Metatype Service
    • OSGi Configuration Admin Service
    • OSGi DTO
    • OSGi Declarative Services
    • OSGi Converter
    • OSGi JPA Service
    • OSGi HTTP Whiteboard
    • OSGi JAX-RS Whiteboard
    • OSGi Transaction Control Service
    • OSGi Push Stream
    • OSGi Promises
  • Equinox OSGi container 3.13.100

  • JPA API 2.1

  • Hibernate 5.2.12

  • Apache Camel 2.23.1, including Camel RabbitMQ component

  • Apache Felix Http Jetty 4.0.6

  • IBM Watson SDK 6.14.2, including Speech to Text and Tone Analyzer

  • Message broker:

    • Mono deployment: RabbitMQ 3.6.16
    • Kubernetes deployment: RabbitMQ 3.7
  • Relational database:

    • Mono deployment: PostgreSQL 10.7
    • Kubernetes deployment: CockroachDB 2.1.6