trenzz-summarize

Inspiration

I have been thinking of an application since long that provides the political mood of country based on twitter trends. This has been first step, main intent of this effort to explore the basic capabilities of the quarkus framework like exposing rest service, invoking the rest service, and connecting database. All this along with ability to create a native image and execute that exploits the advantages of GraalVM.

What it does

The usecase here is application that captures the Top 10 Trending keywords from Twitter for given place and displays that as a graphs on UI. HTML, Java Script and D3 Framework are used for UI rendering.

How I built it

Trenzz Summarize is an application that is developed to showcase the capabilities of the Quarkus. The application uses Quarkus, Quarkus Resteasy, Quarkus Rest Client, Quarkus Scheduler, Quarkus Hibernate Orm, and Quarkus JDBC H2 (as embedded In-memory database).

Challenges I ran into

I was using the dockers for the first time and I ran into multiple challenges with hardware limitation with my old desktop, corporate network restrictions on my office latptop and above all running dockers on Windows. GITHUB actions came to rescue, with it I was able to build docker images easily and I was able to run the container on AWS Fargate, Openshift and Heroku.

I was able to build native image of the application but could not make it working as I was using H2 embedded in-memory database and currently Quarkus does not support executing H2 embedded in-memory database.

Accomplishments that I'm proud of

I was able to build native image of the application. I was able to deploy the application in to multiple different cloud environment.

What I learned

Apart from the basics around Quarkus, I also learned dockers, github actions and resteasy client proxy during this effort.

What's next for trenzz-summarize

I need to fix the UI rendering issue to refresh in the ascending order of time.

As next step I need to run the application as native image by splitting the database and I need to build the intelligence that translates the trend keywords in to political mood.

Built With

• h2
• quarkus
• quarkus-hibernate-orm
• quarkus-rest-client
• quarkus-scheduler