Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Java stored procedure performance dilemma

Re: Java stored procedure performance dilemma

From: rob burton <robburton_at_totalise.co.uk>
Date: 21 Oct 2002 03:38:12 -0700
Message-ID: <feb77408.0210210238.43578849@posting.google.com>


colinandkaren_at_lycos.co.uk (Colin McGuire) wrote in message news:<ab6cea37.0210161156.423d90ba_at_posting.google.com>...
> I am involved with small team of programmers that have just finished
> writing a suite of Java routines that perform some grunt-job string
> manipulation/processing :- the algorithms implemented as Java stored
> procedures. Unfortunately, at this late stage in the play, performance
> has become an issue and we realise we have probably made an incorrect
> technical decision early on. Our reasons for initially choosing Java
> as the development environment were the familiar ones but platform
> independence should not have been one of the reasons as our code is
> intimately tied to Oracle. During a recent technical meeting it was
> mentioned one solution might be to rewrite (ie translate) our library
> routines from Java to Pro/C++, the latter I believe shipped with all
> flavours of Oracle and thus would run on every machine Oracle ran on,
> compile to fast native code etc. We have decided on having a think
> about this approach, weigh up the pro's and con's, and discuss again
> from an informed perspective at the next meeting.
>
> I would be interested in opinions from interested parties in the
> forum, other alternatives to our dilemma etc.
> Thank you
> Colin

Have you tested the performance difference between the java library calls in and out of the database. We had a severe performance problem with an RSA cryptography library that was loaded into the database and and then called from java stored procedures. The procedure was encrypting/decrypting strings of a few K in length using strong public key encryption, and taking about a minute to do this. In the end we moved the library outside of the database without changing the stored procedures greatly and then replaced the crytographic library calls with RMI calls from the database and the performance difference was getting on for a hundred times faster (about a second compared to a minute previously). This was on 8.1.6 (very early days for java in the Oracle server, and I don't think the native compilation option was available then.)

Rob.. Received on Mon Oct 21 2002 - 05:38:12 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US