Lucas Lech

Hey

Your testing method is a bit naive. You should most likely profile each iteration method in isolation for at least few minutes
to reach any conclusions. Actually I did something similar when testing mahout claims they had "fast" collections.
My findings were that the good old simple for was the fastest of the lot. I don't have the results handy but I'll try to find them.

Regards,
Lucas

Fred Hamilton - I can't agree it's the best way to go, tweaking existing code to add the 'statistics gathering' by hand - I thought that's what the tools were for ? But hey, I might be wrong ;)

256 heap is one thing - very small
the way your application behaves might also be an issue - allocating many short-lived objects in a
short amount of time ? Try experimenting the GC options for sizing the space allocated to object generations -
decent read on that here: http://java.sun.com/docs/hotspot/gc5.0/gc_tuning_5.html

Rakesh

The suggested route would be a plsql trigger invoking a java program (plsql wrappers around classes, research more on google, I'm just writing from the top of my head) which in turn would post the ids range onto a jms destination.

Messaging itself should not bring a performance penalty, especially if you're sending metadata & not the data itself. Since you're using a "simple jar" I would suggest using an opensource, standalone jms provider, like activemq (http://activemq.apache.org/) which is fairly easy to setup & tune.

YourKit profiler is my tool of choice - http://www.yourkit.com/

I would go with JMS and a pool of readers ;)

Maybe you could come up with a "hacked" setup as follows:

plsql would send a jms message with a range of row ids (start - end) to a dedicated jms destination (calling an embedded java program ?)
your "serverside" queue listener would pick that information up, and act accordingly - if the order of processing wasn't the issue you could
split the work, even between nodes

LMAO ;)

the questions I got seemed to be sorted by topic so it started with session management, jsps, jstl, custom tags, security, etc., and patterns at the very end

it's kind of hard to really recall the exact questions - the patterns were an easy win, extremely easy questions, read the chapter from the HeadFirst book twice
and you'll be sorted, same with security - moderately hard questions, rest of the stuff, well, similar to what you get in JWebPlus really


study & best of luck ;)

Lucas

they have a lot of coffee and then they have some more

Hi All

I managed to clear this cert today ;) Just for other students benefit, some info on the preparation:

Head First Servlets & JSP, Second Edition - great book, highly recommended

Enthuware JWebPlus mock is a good investment as well (http://www.enthuware.com/jwebplus/index.html)

Whizlabs mocks are a waste of time IMHO, (brief and icomplete answers, kind of feels like an unfinished/rushed-out product)

I would also recommend checking with the specs while doing the mocks, helps a lot.


Lucas

Bear Bibeault wrote:
Even though I wrote a book on Prototype, I no longer recommend it. I've just run into too many problems with it.

would that mean you're now in the jQuery camp or support some other 'player' ? ;)

oh, try xstream for converting objects to json strings as well ;)

hi

first of all you have no quotes, so this is not a correct array, what you need to return would look like
that:

[ 'str1' , 'str2', ... , 'strn']

secondly, you really really should think about using prototypejs or jquery to do such stuff - will make your
life easier

since prototypejs is my favourite, I'll recommend a link ;)) http://www.prototypejs.org/
Cheers,
Lucas

Hi All

Writing to let you know I have posted a refurbished plugin aimed
specifically at EJB3 + Struts2 (xwork2) + JBoss integration.

Feel free to download & test: plugin page on google code

Cheers,
Lucas
[ October 31, 2008: Message edited by: Lucas Lech ]