Ask questions and see you at January, 8th , 6.PM. CET: http://www.ustream.tv/channel/adambien
Also checkout recent episode:
Ask questions and see you at January, 8th , 6.PM. CET: http://www.ustream.tv/channel/adambien
Also checkout recent episode:
Hi Adam.
What would you prefer for a Server push implementation: websocket or SSE? Are These Protocols suitable for long living client to server Connections? E.g. for near realtime information update (stock update for products).
Thank you and Best regards from Hamburg.
Bastian
Hi Adam,
I want to serve a socket from a Java EE container and wonder if it’s a good idea to start accepting client connections in a container managed thread, e.g. using ManagedExecutorService
.
It seems reasonable for me, but on the other hand I’ve found out, that serving a socket is prohibited in EJBs, "Because if an enterprise bean is listening on a socket, it can't be passivated -- it must always be available.“
I’ve looked into JCA as an alternative to communicate with a „socket accepting application", but it’s way more complicated.
An another alternative I've found would be to start listening to the socket in a Servlet, but it seems to be a misuse, because it's for HTTP communication.
I also need outbound communication, not only inbound.
Concretely, I want to write an XMPP (Chat) Server which listens on port 5222 for client connections, but use all the Java EE goodies (CDI, JPA, JAX-RS, …). XMPP allows for alternative connection methods which are HTTP and WebSockets. Both could be simply realized with JavaEE APIs, but not the TCP/socket connections.
My idea is to use Netty for managing socket connections and thought I could pass a ManagedExecutorService
to Netty’s EventLoopGroup
:
new NioEventLoopGroup(1, managedExecutorService)
What's the best architectural approach? JCA, Servlets or a startup bean with ManagedExecutorService
(or similar thread handling)?
Thanks, and a happy new year!
Christian
Happy New Year, Adam!
Just one question about Java EE 8 and JAX-RS + Bean Validation + CDI.
How to get work CDI injection within custom validators? I prepared simple example and seems like everything is ok with it, but it is failing when it tries to inject BookService into the custom ConstraintValidator. How to configure the correct behavior of JAX-RS + Bean Validation + CDI? Obviously, I can get every dependency I needed via CDI.current().select(SomeClass.class).get() and it works fine, but it looks dirty and quirky. This example doesn't work on Glassfish 4,5 and Payara 4,5, but it works perfect on openliberty server and on wildfly. Anyway, I need to run it on payara 5 somehow. It also works on TomEE, but only when Validator specified explicitly on custom validation annotation, but it doesn't when specified at META-INF/services/javax.validation.ConstraintValidator
Thanks!
I created the workaround here, it works, but there should be easier solution https://github.com/afrunt/examples/tree/master/misc/jaxrs-bean-validation-cdi-custom-validator
sources of original example are here https://github.com/afrunt/examples/tree/master/java-ee-8-examples/bean-validation-custom-validator
@Target({FIELD})
@Retention(RUNTIME)
@Constraint(validatedBy = {})
@Documented
public @interface JavaEE8Book {
String message() default "{com.afrunt.example.validation.constraints.javaee8book.message}";
Class<?>[] groups() default {};
Class<? extends Payload>[] payload() default {};
}
@Named
@ApplicationScoped
public class JavaEE8BookValidator implements ConstraintValidator<JavaEE8Book, String> {
@Inject
BookService bookService;
@Override
public boolean isValid(String name, ConstraintValidatorContext context) {
//return CDI.current().select(BookService.class).get().isValidBookName(name);
return bookService.isValidBookName(name);
}
}
Workaround for Glassfish/Payara 5
@Provider
public class CDIProviderResolver implements ContextResolver<ValidationConfig> {
@Context
ResourceContext context;
@Override
public ValidationConfig getContext(Class<?> type) {
return new ValidationConfig().constraintValidatorFactory(new CDIConstraintValidatorFactory(context));
}
public static class CDIConstraintValidatorFactory implements ConstraintValidatorFactory {
private ResourceContext context;
CDIConstraintValidatorFactory(ResourceContext context) {
this.context = context;
}
@Override
public <T extends ConstraintValidator<?, ?>> T getInstance(Class<T> key) {
try {
return CDI.current().select(key).get();
} catch (Exception e) {
return context.getResource(key);
}
}
@Override
public void releaseInstance(ConstraintValidator<?, ?> instance) {
}
}
}
Hi Adam,
i really like your really thin war files which are fast to deploy and I'm currently trying to have a really small war for my small private project too. The setup is based on your java ee maven archetype with the glassfish docker image.
To keep the war small, i tried to work with the built in java.util.logging instead of slf4j which i used before. By default, the logger logs into the server log of glassfish. But I'd like to have a single log file only for the log entries of my application. I tried to add a logging.properties file to the project to define the FileHandler but this file then overrides all the properties of the server. How do you achieve to keep the server config untouched and add logging config? How do you do that in your projects?
In addition i would like to send the log entries via filebeat to ELK. Would you add filebeat to the docker image or would you have a dedicated filebeat docker container with access to the application log files? Or another way?
Thanks in advance
Stefan
Hi Adam,
1-When you deploy an ear or war to a client server, is there a good mecanism in order to obfuscate the code so the client cant reverse engineer it.
2- If we obfuscate the code , do the application server still be able to execute it properly?
3- What is the best way to add a license to your JEE application when you want to deploy it in a client server.
thanks.
Hello!
Thanks for the videos on youtube and the online courses on Vimeo! Really appreciate it.
Question :
I know your view on external libraries and I do understand the reason behind avoiding them.
However, Since I really started using streams, there are times when you are using e.g. IntStream and then getting a value from a structure. (they don't provide Iterator or list output -- don't ask why :) )
Now further down the stream I might also want the integer generated by the stream but normally you cannot as once you do like a map(i -> ds.getObjectAt(i)) the i is not available anymore!
Here I have been saved many a times by Pair as I can have a Object with both values in the Stream.
What do you do in your projects? Create your own each time? or is this good candidate for using Apache Commons?
Hi Adam,
are you open for companies sponsoring your airhacks-webshows?
Hi Adam,
I missed the previous gist. If you want an update on my old issue, I found that for my case the simplest solution to work with the RequestScoped bean which was holding the state was adding a AroundInvoke interceptor to extract the state from its holder bean and assign it to a field in the stateless beans. I didn't like passing the state around as an argument to all the methods of all stateless beans.
Anyhow, my questions:
public int add(int a, int b)
...), does it matter if I put each method in a corresponding stateless bean or all 4 in one? I know that in the first case there will be less of each bean in the pool and in the latter more of the single one. Should I just do what makes the most sense? Is it harmful to have a "god object" stateless bean or the opposite?Also thank you for very good advice ;)