I have an application which uses a lot of Java actions involving batch retrieves and IBatch processing, and it runs three or four times slower in 2.5 than it did in 188.8.131.52. I am trying to discern the reason for this. As far as I can see the only differences are: in 2.5 we pass the context when we call IMendixObject.setValue, etc. In 184.108.40.206 we didn't pass any context. in 220.127.116.11 I used to kick the user out during long processes. I can't do this in 2.5 because I can't get hold of an ISession object. there is the odd System.out.println kicking about and these no longer display in the console. Are there any performance recommendations on conversion to 2.5 associated with these volume processing techniques please?
It's kinda hard to discern what's going on exactly without knowing what kind of application you're running. You can try running the microflow debugger and stepping through the activities, this will tell you how much time each activity uses, which might shed some light on what's causing the problems.
As to the differences you've pointed out:
Passing the context shouldn't make
any difference, as far as I can tell.
You can get the ISession object
anywhere by calling getSession on the
System.out's won't work anymore
because the new logging system
works via HTTP polling. We strongly suggest using the logger
interfaces that we've provided to give feedback during execution.
You can create a new loggers by
simply providing a new string name to