[wellylug] Java 32 Bit/ Java 64 Bit Performance

[Gerald.Roehrbein]

Hello Michael,

the records are small.
Database time is just 4 minutes in 45 minutes. The Java "computation"
takes 41 minutes.

The software fetches only a few bytes. Most records of the result sets
are one number rows.

If the software would take 4 minutes than it would make sense to
increase the arraysize (thats the number of records of the result set
Oracle put in one ethernet package).

Database and network do not matter.

It's just bad Java and not simple to change (I do not have source code).

My recommendation was to use a Java profiler to find the container class
and method which takes most time. The developers said that they will
install a profiler and identify the bad code but it will take a few
month to fix.



kind regards
Gerald 


On Tue, 2009-04-28 at 21:37 +1200, michael at livia.co.nz wrote:
> >a) 32 Bit Linux + 32 Bit Java memory model gave best performance
> >especially with memory allocation. Test program took 45 minutes.
> 
> >The reason for the difference was the duration of some types of memory
> >allocations. The application reads 70.000 records from a database into a
> >java clients memory.
> 
> How wide are these records that reading 70k of them takes 45 minutes on
> any sort of modern system?
> 
> Michael.