interesting - yes the only downside I have noted with 64bit "all the way" is *sometimes* a tad more memory is used for the same thing (generally due to pointers being twice the size) but its not huge (20% more memory for small programs, less noticably in the large). Java 6 JVMs (Sun and Apple) both have options in hotspot for helping reduce this for apps that have a heap <32Gig (above that, you won't notice anyway I guess).
On Wed, Oct 7, 2009 at 6:50 PM, James Owen <[email protected]> wrote: > Greetings: > I've been running some benchmarks on some other systems recently using the > WaltzDB-200 benchmark. (I have to write the data / rules generator for each > vendor and Drools will probably be next after Jess and CLIPS. I've already > done OPSJ and TECH.) Using Java 6 / 64-bit system over the Java 5 system > dramatically improves performance for the two that I have tried so far and > there is no reason to believe that it wouldn't dramatically improve Drools > as well. The other thing that really improves performance is moving to a > 64-bit systems. Most desktop units (and laptops) are 32-bit OS, not 64-bit. > The final nail in the coffin was running DDR3 memory rather than DDR2 > memory. All together I got 10 times the performance or better. When you > put all of this together, you have a smoking machine. > So, my vote is for Java 6, 64-bit OS and DDR 3 for any machine: Server, > Desktop, laptop, notebook, whatever... > SDG > James Owen > Founder October Rules Fest > Senior Consultant / Architect KBSC > http://www.kbsc.com > http://www.OctoberRulesFest.org > Twitter: OctRulesFest > Blogs: > http://JavaRules.blogspot.com [Rulebased Systems Blog] > http://ORF2009.blogspot.com [October Rules Fest Blog] > http://exscg.blogspot.com/ [Expert Systems Consulting Group Blog] > "If I have seen a little further it is by standing on the shoulders of > giants." > Sir Isaac Newton in a letter to Robert Hooke, 5 Feb 1676 > Come to October Rules Fest and stand on the shoulders of the Giants of the > industry; if only for a week. > > > On Oct 7, 2009, at 2:38 AM, Michael Neale wrote: > > yeah its a pain - even if we have to target 1.5 (which is entirely > reasonable for the next few releases) we can probably *require* that > it is built with JDK6, using whatever dirty tricks there are to make > the bytecode work with 1.5. > > Apparently the EOL is for the Sun JVM, the IBM and other ones have a > longer life, it seems (and in many cases, the EOL of support on the > core JVM is ignored by users, more so then the app servers or > libraries or applications they run on top ! I have no idea why it is > that way, it just is. Probably like how some places still have Windows > 2000 or NT around !). > > > On Wed, Oct 7, 2009 at 4:44 AM, Geoffrey De Smet > <[email protected]> wrote: > > To compile trunk with Java 6, for now you have to use "-source 1.5". > > That doesn't work, as the maven compiler plugin is already set to use > > source 1.5 in the main pom and I can't compile with jdk 6. It can only > > be compiled with jdk 5 it seems, not jdk 6. > > With kind regards, > > Geoffrey De Smet > > > Edson Tirelli schreef: > > Geoffrey, > > Java 6 API change was not source backward compatible with Java 5. > > Drools 5.1m1 is supposed to compile only with Java 5. > > We will probably move to Java 6 for the next milestone, since Java 5 > > is EOL this month anyway, but you need to rollback your change meanwhile. > > To compile trunk with Java 6, for now you have to use "-source 1.5". > > Edson > > > 2009/10/6 Geoffrey De Smet <[email protected] > > <mailto:[email protected]>> > > I've fixed it on trunk with this commit: > > "compilation error on jdk 6 (not on jdk 5) because the ExectutorService > > interface in java 6 follows the Joshua Bloch PECS pattern (so > > Collection<? extends Callable...> instead of just > > Collection<Callable...>)" > > See > > > http://fisheye.jboss.org/browse/JBossRules/trunk/drools-core/src/main/java/org/drools/concurrent/ExternalExecutorService.java?r1=29459&r2=29541 > > > <http://fisheye.jboss.org/browse/JBossRules/trunk/drools-core/src/main/java/org/drools/concurrent/ExternalExecutorService.java?r1=29459&r2=29541> > > With kind regards, > > Geoffrey De Smet > > > Geoffrey De Smet schreef: > > > When I try "mvn clean install -DskipTests", I get this (java version > > > "1.6.0_16"): > > > > > > [INFO] > > > > > ------------------------------------------------------------------------ > > > [ERROR] BUILD FAILURE > > > [INFO] > > > > > ------------------------------------------------------------------------ > > > [INFO] Compilation failure > > > > > > /home/ge0ffrey/projects/jboss/drools/drools-core/src/main/java/org/drools/concurrent/ExternalExecutorService.java:[31,7] > > > org.drools.concurrent.ExternalExecutorService is not abstract and > > does > > > not override abstract method <T>invokeAny(java.util.Collection<? > > extends > > > > > java.util.concurrent.Callable<T>>,long,java.util.concurrent.TimeUnit) in > > > java.util.concurrent.ExecutorService > > > > > > > > > > > > > > > /home/ge0ffrey/projects/jboss/drools/drools-core/src/main/java/org/drools/concurrent/ExternalExecutorService.java:[31,7] > > > org.drools.concurrent.ExternalExecutorService is not abstract and > > does > > > not override abstract method <T>invokeAny(java.util.Collection<? > > extends > > > > > java.util.concurrent.Callable<T>>,long,java.util.concurrent.TimeUnit) in > > > java.util.concurrent.ExecutorService > > > > > > > > _______________________________________________ > > rules-dev mailing list > > [email protected] <mailto:[email protected]> > > https://lists.jboss.org/mailman/listinfo/rules-dev > > > > > -- > > Edson Tirelli > > JBoss Drools Core Development > > JBoss by Red Hat @ www.jboss.com <http://www.jboss.com> > > > ------------------------------------------------------------------------ > > _______________________________________________ > > rules-dev mailing list > > [email protected] > > https://lists.jboss.org/mailman/listinfo/rules-dev > > _______________________________________________ > > rules-dev mailing list > > [email protected] > > https://lists.jboss.org/mailman/listinfo/rules-dev > > > > > -- > Michael D Neale > home: www.michaelneale.net > blog: michaelneale.blogspot.com > > _______________________________________________ > rules-dev mailing list > [email protected] > https://lists.jboss.org/mailman/listinfo/rules-dev > > > _______________________________________________ > rules-dev mailing list > [email protected] > https://lists.jboss.org/mailman/listinfo/rules-dev > > -- Michael D Neale home: www.michaelneale.net blog: michaelneale.blogspot.com _______________________________________________ rules-dev mailing list [email protected] https://lists.jboss.org/mailman/listinfo/rules-dev
