[rules-users] Drools vs OPSJ

Edson Tirelli ed.tirelli at gmail.com
Thu Apr 8 09:19:38 EDT 2010


    Hey Benoit,

    That is why we always tell people to ignore the classic benchmarks
(manners, waltz, etc) when comparing rules engines and use their own use
cases to test. There are so many different optimizations implemented in
modern rules engines that looking at one single aspect of them through an
academic benchmark "lens" will never give you the whole picture, an so what
you need is your own use case. That is also why it is so hard to get vendors
to agree on new benchmarks, because each of them will try to come up with a
benchmark that emphasizes his strengths. Although these academic benchmarks
were helpful in the past, I believe nowadays they don't help anyone anymore.

    So, you did the right thing. Look at your use case, test the engines and
find what is best for you. You are not the only one finding such results for
real use cases.

    Also, performance is just one variable in the equation. Most rules
engines will be fast enough for your needs, but what about rules authoring?
declarative power? integration with your architecture? maintenance? support?
cost?

    Good luck and obviously count on us to help you with any difficulty you
face.

    Edson

2010/4/8 Bertrand Grottier <b_grottier at hotmail.com>

>  Hello,
>
> At the present time, I am evaluating Drools vs OPSJ.
>
> Before I started, I had a look at the following tests:
> http://blogs.illation.com.au/2009/09/waltz-benchmark-performance-opsj-vs-drools/.
> Based on these tests, I was expecting better performances from OPSJ over
> Drools regarding the rule firing time.
> However, I obtained opposite results on a basic test. It consists in a sort
> of "join clause": I have a few products and many potentiel prices (10^4) in
> the knowledge session. The aim is to "match" the appropriate prices to each
> products.
>
> My question are the followings:
> - Does someone obtained similar results ?
> - Are there situations where Drools outperforms OPSJ ? (apparently yes).
>
> Thank you in advance.
>
> Regards,
>
> Benoît
>
> ------------------------------
> Commander un cadeau en ligne discrètement? Voir la solution offerte par
> Internet Explorer8 <http://clk.atdmt.com/FRM/go/207186971/direct/01/>
>
> _______________________________________________
> rules-users mailing list
> rules-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users
>
>


-- 
 Edson Tirelli
 JBoss Drools Core Development
 JBoss by Red Hat @ www.jboss.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/rules-users/attachments/20100408/8e5ff7c5/attachment.html 


More information about the rules-users mailing list