20thAirhacksQ&A.md

20thAirhacksQ&A.md

Ask questions and see you at November, 2nd at 6.PM. CET: http://www.ustream.tv/channel/adambien Also see archives: airhacks.tv

Hi Adam,

We are building a Java EE web application which makes moderately to heavy use of DB resources.
A lot of the DB table is configuration stuff.
At the moment we are only able to have around 250 concurrent users in an 8 core 16 GB of RAM Server machine.
We have not done any JVM tuning, nor used batch processing with JPA, so far just plain JPA relationships with defaults.

My question is:
In your projects how many users do you support in a server with similar characteristics?
I know that this is difficult and changes from app to app but I would like to have some numbers on an average boring Java EE application.

How many concurrent users has it been able to support without/before any optimisation performed?

Thanks.

Hello again 😆
What is the best api to implement SSE? REST+Ajax or Websockets ?

Thank you so much 😄

Hello Adam

You've previously show neat tricks, like marshalling/unmarshalling json to pojo's.

Staying in the json realm :

Do you have an easy way to validate input/output (@XmlRootElements), when calling method annotated like this:

    @POST
    @Path("/")
    @Consumes(MediaType.APPLICATION_JSON)
    @Produces(MediaType.APPLICATION_JSON)
    public Out method(In in) {
...

Having looked at examples, I hope I've missed a trick - you can't have to jump through all those hoops.

It would be nice to know that when the method is invoked, all required fields are populated (doing null pointer checks, feels like you're doing it wrong).

See You