93rdAirhacksQ&A.md

93rdAirhacksQ&A.md

Ask questions and see you at December, 2nd, 8.PM. CET: youtube.com/c/bienadam/live

Also checkout recent episode:

Please keep the questions Jakarta EE-stic. Means: as short and as concise as only possible. Feel free to ask several, shorter questions. Upcoming airhacks.tv events are also going to be announced at meetup.com/airhacks

Hi @AdamBien,

could you please explain me the differences between the following three annotations:

• @singleton
• @stateless
• @ApplicationScoped

If @singleton means that every client that uses that bean receives the same (and only) instance defined in the DI container why was it necessary to define the annotation @ApplicationScoped? In what are @singleton and @ApplicationScoped different? Also, I find it very hard sometimes to decide if my beans should be singletons or stateless because to me they sound the same. I read the documentation several times but somehow the explanation did not resonate to me very much. Could you please tell me how these annotations differ from each other?

Thanks in advance!

Hi @AdamBien,

I very often make use of an EntityManagerProducer class and annotate the method that returns the EntityManager with @produces so that I can inject the produced bean in other beans. This way I can use constructor injection and simplify my life when writing unit and integration tests because I am free to define whatever persistence unit I want.

I have read several times, however, that the EntityManager is not thread safe and I was wondering two things:

• what are the implications of the entity manager not being thread safe? What kind of issues can I incurr into if I inject it into several other beans and not paying attention to thread safety?
• how can I correctly produce the entity manager in a way that it is thread safe?

Thanks again.

When trying to use your template aws-quarkus-lambda-cdk-plain, it took several hours to get it to deploy due to a bootstrap issue.

I finally found the following worked to properly bootstrap the app:

export CDK_NEW_BOOTSTRAP=1
cdk bootstrap aws://ACCOUNT-NUMBER/REGION

Evidently the "modern" template is required for bootstrap and the above command configures it correctly.

Can you explain how to properly setup aws cdk/lambda via bootstrap and config and if an IAM user is required?

Thanks much

Hello sir,

What is your approach to development, test, production stages when it comes to persistence.xml file, how do you use the schema creation feature, and how to protect production from deploying some artifact with a wrongly attributed persistence.xml where javax.persistence.schema-generation.database.action is set to drop-and-create option?

Thank you very much.

Hey! What do you think of the one container one process approach? And multithreading is taken over by the orchestration system through the creation of the required number of containers. Is this approach good for Java applications?

The microprofile fault tolerance api will allow us to limit the number of concurrent calls to our business method with Bulkhead annotation. Is it possible to use the same annotation for the microprofile rest client method in interface definition?

Graviton and Ampere will reduce the cost of cloud services?

Hi Adam, thanks for the great Q&A sessions.

Since your Devflex session "Hey Enterprise EJB Developers Now Is The Time To Go Serverless" I can't get this topic Serverless Computing out of my head. I have already migrated parts of our modular monolith (Jakarta EE 8) to Quarkus as part of a POC and tried it out as an Azure Function. It was very exciting and interesting.

I would be interested if it is realistic to migrate a modular monolith completely to Serverless. Do you have any experience with such projects?

Can you possibly show an example how the orchestration of such a serverless architecture in production can look like? Communication between the functions, 0-downtime deployment etc.

Greetings
Denis