7:47 p.m.
Progress has recently ripped off the Drools community and has offered
nothing in return, not even a curtesy nod of accreditation for our hard
work.. 30% of that 49mill USD would go a long way to helping us invest
in our tooling which would benefit all of us :) and maybe the Drools
team can have it's next team meeting in Hawaii and pay for all the
community members to come too. Cocktails anyone? :)
http://blog.athico.com/2010/09/progress-buy-bridge-in-brooklyn-savvion.html
I'm just having a little bit of fun, at their expense, which is well
deserved for such shockingly bad behaviour. So I would urge anyone else
to join in and have some fun too :) Do your own blogs, polls etc to
bring attention to this.
http://blog.athico.com/2010/09/breaking-news-have-your-vote-on-what.html
And please vote up at Digg and DZone to help raise awareness:
http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
<http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
http://digg.com/news/technology/drools_progress_buy_a_bridge_in_brooklyn_...
Being slightly less tongue in check, this is just the reality of open
source, I've been doing this a long time now and I know you just have to
live with it, take the lumps with the smooth - as long as you get more
smooth than lumps, it's still worthwhile :) At it's best it brings
people together who collaborate and help each other and enrich the
open-source eco system. At it's worst you have organisations such as
Progress executing on predatory amoral tactics. The debate on whether
all corporations are ultimately operating on degrees of amorality is out
of scope for today :) As each year goes buy you start to appreciate Red
Hat more for the sterling work it does in open source. While it would be
nice if Progress got involved and contributed to Drools, the reality is
that this doesn't both me as much as you think it would. It hasn't
lestened the Drools community and technology, we still go from strength
to strength, and I've always been more concerned about what we are doing
and achieving than what someone else may or may not be doing with
regards to taking advantage of our work.
Ultimately this is actually quite flattering and re-affirms the strength
of our technology and the directions we are going. I'd rather have an
organisation using Drools and moving it one step closing to being a
defacto standard, than they partner with someone else or develop yet
another rule engine splintering the market and confusing users. "Drools
Everywhere" has alwas been my moto :)
However that said, anyone who knows me, knows I have a mischievous sense
of humour and if Progress and Savvion aren't even going to at the very
least accredit us then they are fair game and surely myself and the
Drools community deserve a little fun at their expense :)
Mark - The Mischievous Lead
9:46 p.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
I've observed a sizable (25%) performance degradation between the 5.0.1 and 5.1.0
releases. I've attached a sample project that tests the performance of matching
nested objects. (ANd compares direct reference matching and the performance pitfalls of
"from," but that's beside the current point.)
If you switch the pom.xml from using 5.0.1 to 5.1.0 for the drools dependencies you'll
see 25% longer execution times on the tests. (mvn test)
Here's the test output:
5.0.1
reference.drl Count: 2000reference.drl Time: 267msreference.drl Time per element:
0.1335msBAR Duplicates: 780FOO Duplicates: 880
reference.drl Count: 20000reference.drl Time: 1249msreference.drl Time per element:
0.06245msBAR Duplicates: 7702FOO Duplicates: 8040
from.drl Count: 200from.drl Time: 1139msfrom.drl Time per element: 5.695msBAR Duplicates:
112FOO Duplicates: 102
reference.drl Count: 200reference.drl Time: 5msreference.drl Time per element: 0.025msBAR
Duplicates: 86FOO Duplicates: 60
5.1.0
reference.drl Count: 2000reference.drl Time: 300msreference.drl Time per element:
0.15msBAR Duplicates: 788FOO Duplicates: 820
reference.drl Count: 20000reference.drl Time: 1564msreference.drl Time per element:
0.0782msBAR Duplicates: 8142FOO Duplicates: 7960
from.drl Count: 200from.drl Time: 3543msfrom.drl Time per element: 17.715msBAR Duplicates:
68FOO Duplicates: 90
reference.drl Count: 200reference.drl Time: 13msreference.drl Time per element: 0.065msBAR
Duplicates: 84FOO Duplicates: 74
On the most taxing test (20k objects) 5.0.1 took 1249ms while 5.1.0 took 1564ms, and for
larger tests the effect is more pronounced. This is primarily a test of == on object
references.
GreG
9:49 p.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
Thanks we'll look into it, now if you only gave that to us before 5.1
final ;)
In general we've had some great performance improvements on systems with
large number of facts due to "differential update".
Will be interesting to see where this slow down point is.
Thanks
Mark
On 03/09/2010 03:46, Greg Barton wrote:
I've observed a sizable (25%) performance degradation between the
5.0.1 and 5.1.0 releases. I've attached a sample project that tests
the performance of matching nested objects. (ANd compares direct
reference matching and the performance pitfalls of "from," but that's
beside the current point.)
If you switch the pom.xml from using 5.0.1 to 5.1.0 for the drools
dependencies you'll see 25% longer execution times on the tests. (mvn
test)
Here's the test output:
5.0.1
reference.drl Count: 2000
reference.drl Time: 267ms
reference.drl Time per element: 0.1335ms
BAR Duplicates: 780
FOO Duplicates: 880
reference.drl Count: 20000
reference.drl Time: 1249ms
reference.drl Time per element: 0.06245ms
BAR Duplicates: 7702
FOO Duplicates: 8040
from.drl Count: 200
from.drl Time: 1139ms
from.drl Time per element: 5.695ms
BAR Duplicates: 112
FOO Duplicates: 102
reference.drl Count: 200
reference.drl Time: 5ms
reference.drl Time per element: 0.025ms
BAR Duplicates: 86
FOO Duplicates: 60
5.1.0
reference.drl Count: 2000
reference.drl Time: 300ms
reference.drl Time per element: 0.15ms
BAR Duplicates: 788
FOO Duplicates: 820
reference.drl Count: 20000
reference.drl Time: 1564ms
reference.drl Time per element: 0.0782ms
BAR Duplicates: 8142
FOO Duplicates: 7960
from.drl Count: 200
from.drl Time: 3543ms
from.drl Time per element: 17.715ms
BAR Duplicates: 68
FOO Duplicates: 90
reference.drl Count: 200
reference.drl Time: 13ms
reference.drl Time per element: 0.065ms
BAR Duplicates: 84
FOO Duplicates: 74
On the most taxing test (20k objects) 5.0.1 took 1249ms while 5.1.0
took 1564ms, and for larger tests the effect is more pronounced. This
is primarily a test of == on object references.
GreG
_______________________________________________
rules-dev mailing list
rules-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
9:51 p.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
I think the attachment is missing, maybe use jira and attach it there?
Mark
On 03/09/2010 03:46, Greg Barton wrote:
I've observed a sizable (25%) performance degradation between the
5.0.1 and 5.1.0 releases. I've attached a sample project that tests
the performance of matching nested objects. (ANd compares direct
reference matching and the performance pitfalls of "from," but that's
beside the current point.)
If you switch the pom.xml from using 5.0.1 to 5.1.0 for the drools
dependencies you'll see 25% longer execution times on the tests. (mvn
test)
Here's the test output:
5.0.1
reference.drl Count: 2000
reference.drl Time: 267ms
reference.drl Time per element: 0.1335ms
BAR Duplicates: 780
FOO Duplicates: 880
reference.drl Count: 20000
reference.drl Time: 1249ms
reference.drl Time per element: 0.06245ms
BAR Duplicates: 7702
FOO Duplicates: 8040
from.drl Count: 200
from.drl Time: 1139ms
from.drl Time per element: 5.695ms
BAR Duplicates: 112
FOO Duplicates: 102
reference.drl Count: 200
reference.drl Time: 5ms
reference.drl Time per element: 0.025ms
BAR Duplicates: 86
FOO Duplicates: 60
5.1.0
reference.drl Count: 2000
reference.drl Time: 300ms
reference.drl Time per element: 0.15ms
BAR Duplicates: 788
FOO Duplicates: 820
reference.drl Count: 20000
reference.drl Time: 1564ms
reference.drl Time per element: 0.0782ms
BAR Duplicates: 8142
FOO Duplicates: 7960
from.drl Count: 200
from.drl Time: 3543ms
from.drl Time per element: 17.715ms
BAR Duplicates: 68
FOO Duplicates: 90
reference.drl Count: 200
reference.drl Time: 13ms
reference.drl Time per element: 0.065ms
BAR Duplicates: 84
FOO Duplicates: 74
On the most taxing test (20k objects) 5.0.1 took 1249ms while 5.1.0
took 1564ms, and for larger tests the effect is more pronounced. This
is primarily a test of == on object references.
GreG
_______________________________________________
rules-dev mailing list
rules-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
10:33 p.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
would be interesting to see memory usage as well. I know its the poor red
headed cousin to execution time performance, but it still matters !
On Fri, Sep 3, 2010 at 12:46 PM, Greg Barton <greg_barton(a)yahoo.com> wrote:
I've observed a sizable (25%) performance degradation between the
5.0.1 and
5.1.0 releases. I've attached a sample project that tests the performance
of matching nested objects. (ANd compares direct reference matching and the
performance pitfalls of "from," but that's beside the current point.)
If you switch the pom.xml from using 5.0.1 to 5.1.0 for the drools
dependencies you'll see 25% longer execution times on the tests. (mvn test)
Here's the test output:
5.0.1
reference.drl Count: 2000
reference.drl Time: 267ms
reference.drl Time per element: 0.1335ms
BAR Duplicates: 780
FOO Duplicates: 880
reference.drl Count: 20000
reference.drl Time: 1249ms
reference.drl Time per element: 0.06245ms
BAR Duplicates: 7702
FOO Duplicates: 8040
from.drl Count: 200
from.drl Time: 1139ms
from.drl Time per element: 5.695ms
BAR Duplicates: 112
FOO Duplicates: 102
reference.drl Count: 200
reference.drl Time: 5ms
reference.drl Time per element: 0.025ms
BAR Duplicates: 86
FOO Duplicates: 60
5.1.0
reference.drl Count: 2000
reference.drl Time: 300ms
reference.drl Time per element: 0.15ms
BAR Duplicates: 788
FOO Duplicates: 820
reference.drl Count: 20000
reference.drl Time: 1564ms
reference.drl Time per element: 0.0782ms
BAR Duplicates: 8142
FOO Duplicates: 7960
from.drl Count: 200
from.drl Time: 3543ms
from.drl Time per element: 17.715ms
BAR Duplicates: 68
FOO Duplicates: 90
reference.drl Count: 200
reference.drl Time: 13ms
reference.drl Time per element: 0.065ms
BAR Duplicates: 84
FOO Duplicates: 74
On the most taxing test (20k objects) 5.0.1 took 1249ms while 5.1.0
took 1564ms, and for larger tests the effect is more pronounced. This is
primarily a test of == on object references.
GreG
_______________________________________________
rules-dev mailing list
rules-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
--
Michael D Neale
home: www.michaelneale.net
blog: michaelneale.blogspot.com
11:59 p.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
Greg,
Can you check 5.1.1 instead? There was a problem in 5.1.0 that caused
indexes to be disabled in certain situations and also a bug on "exists". In
general, we had the opposite feedback: reduction of memory usage and
performance increase due to reduced GC overhead.
Edson
2010/9/2 Greg Barton <greg_barton(a)yahoo.com>
I've observed a sizable (25%) performance degradation between the
5.0.1 and
5.1.0 releases. I've attached a sample project that tests the performance
of matching nested objects. (ANd compares direct reference matching and the
performance pitfalls of "from," but that's beside the current point.)
If you switch the pom.xml from using 5.0.1 to 5.1.0 for the drools
dependencies you'll see 25% longer execution times on the tests. (mvn test)
Here's the test output:
5.0.1
reference.drl Count: 2000
reference.drl Time: 267ms
reference.drl Time per element: 0.1335ms
BAR Duplicates: 780
FOO Duplicates: 880
reference.drl Count: 20000
reference.drl Time: 1249ms
reference.drl Time per element: 0.06245ms
BAR Duplicates: 7702
FOO Duplicates: 8040
from.drl Count: 200
from.drl Time: 1139ms
from.drl Time per element: 5.695ms
BAR Duplicates: 112
FOO Duplicates: 102
reference.drl Count: 200
reference.drl Time: 5ms
reference.drl Time per element: 0.025ms
BAR Duplicates: 86
FOO Duplicates: 60
5.1.0
reference.drl Count: 2000
reference.drl Time: 300ms
reference.drl Time per element: 0.15ms
BAR Duplicates: 788
FOO Duplicates: 820
reference.drl Count: 20000
reference.drl Time: 1564ms
reference.drl Time per element: 0.0782ms
BAR Duplicates: 8142
FOO Duplicates: 7960
from.drl Count: 200
from.drl Time: 3543ms
from.drl Time per element: 17.715ms
BAR Duplicates: 68
FOO Duplicates: 90
reference.drl Count: 200
reference.drl Time: 13ms
reference.drl Time per element: 0.065ms
BAR Duplicates: 84
FOO Duplicates: 74
On the most taxing test (20k objects) 5.0.1 took 1249ms while 5.1.0
took 1564ms, and for larger tests the effect is more pronounced. This is
primarily a test of == on object references.
GreG
_______________________________________________
rules-dev mailing list
rules-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-dev
--
Edson Tirelli
JBoss Drools Core Development
JBoss by Red Hat @ www.jboss.com
12:23 a.m.
New subject: FYI: Performance degradation from 5.0.1 to 5.1.0
5.1.1 is slightly faster, but still beaten by 5.0.1.
5.1.1:
reference.drl Count: 80000reference.drl Time: 6812msreference.drl Time per element:
0.08515msBAR Duplicates: 32012FOO Duplicates: 31938
5.1.0:
reference.drl Count: 80000reference.drl Time: 7533msreference.drl Time per element:
0.0941625msBAR Duplicates: 31832FOO Duplicates: 32642
5.0.1:
reference.drl Count: 80000reference.drl Time: 4814msreference.drl Time per element:
0.060175msBAR Duplicates: 32494FOO Duplicates: 31578
I tried running 5.2.0.SNAPSHOT but the tests died with the following exception:
org.drools.RuntimeDroolsException: Unable to load dialect
'org.drools.rule.builder.dialect.java.JavaDialectConfiguration:java:org.drools.rule.builder.dialect.java.JavaDialectConfiguration' at
org.drools.compiler.PackageBuilderConfiguration.addDialect(PackageBuilderConfiguration.java:275)
at
org.drools.compiler.PackageBuilderConfiguration.buildDialectConfigurationMap(PackageBuilderConfiguration.java:260)
at
org.drools.compiler.PackageBuilderConfiguration.init(PackageBuilderConfiguration.java:173)
at
org.drools.compiler.PackageBuilderConfiguration.<init>(PackageBuilderConfiguration.java:151)
at org.drools.compiler.PackageBuilder.<init>(PackageBuilder.java:216)
at org.drools.compiler.PackageBuilder.<init>(PackageBuilder.java:147)
at
org.drools.builder.impl.KnowledgeBuilderFactoryServiceImpl.newKnowledgeBuilder(KnowledgeBuilderFactoryServiceImpl.java:34)
at
org.drools.builder.KnowledgeBuilderFactory.newKnowledgeBuilder(KnowledgeBuilderFactory.java:47)
at com.sample.DroolsTest.readKnowledgeBase(DroolsTest.java:66)
at com.sample.DroolsTest.go(DroolsTest.java:31)
at com.sample.SimpleTest.fromTest_200(SimpleTest.java:29)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:637)
Caused by: java.lang.RuntimeException: The Eclipse JDT Core jar is not in the classpath
at
org.drools.rule.builder.dialect.java.JavaDialectConfiguration.setCompiler(JavaDialectConfiguration.java:94)
at
org.drools.rule.builder.dialect.java.JavaDialectConfiguration.init(JavaDialectConfiguration.java:55)
at
org.drools.compiler.PackageBuilderConfiguration.addDialect(PackageBuilderConfiguration.java:271)
... 33 more
... Removed 20 stack frames
--- On Thu, 9/2/10, Edson Tirelli <tirelli(a)post.com> wrote:
From: Edson Tirelli <tirelli(a)post.com>
Subject: Re: [rules-dev] FYI: Performance degradation from 5.0.1 to 5.1.0
To: "Rules Dev List" <rules-dev(a)lists.jboss.org>
Date: Thursday, September 2, 2010, 11:59 PM
Greg,
Can you check 5.1.1 instead? There was a problem in 5.1.0 that caused indexes to be
disabled in certain situations and also a bug on "exists". In general, we had
the opposite feedback: reduction of memory usage and performance increase due to reduced
GC overhead.
Edson
12:42 p.m.
New subject: [rules-users] Shame on Progress and Savvion
We just made #4 on dzone, please keep the votes coming, let's see if
we can make #1
http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
<http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
Mark
On 03/09/2010 01:47, Mark Proctor wrote:
Progress has recently ripped off the Drools community and has offered
nothing in return, not even a curtesy nod of accreditation for our
hard work.. 30% of that 49mill USD would go a long way to helping us
invest in our tooling which would benefit all of us :) and maybe the
Drools team can have it's next team meeting in Hawaii and pay for all
the community members to come too. Cocktails anyone? :)
http://blog.athico.com/2010/09/progress-buy-bridge-in-brooklyn-savvion.html
I'm just having a little bit of fun, at their expense, which is well
deserved for such shockingly bad behaviour. So I would urge anyone
else to join in and have some fun too :) Do your own blogs, polls etc
to bring attention to this.
http://blog.athico.com/2010/09/breaking-news-have-your-vote-on-what.html
And please vote up at Digg and DZone to help raise awareness:
http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
<http://www.dzone.com/links/progress_buy_a_bridge_in_brooklyn_savvion_brms...
http://digg.com/news/technology/drools_progress_buy_a_bridge_in_brooklyn_...
Being slightly less tongue in check, this is just the reality of open
source, I've been doing this a long time now and I know you just have
to live with it, take the lumps with the smooth - as long as you get
more smooth than lumps, it's still worthwhile :) At it's best it
brings people together who collaborate and help each other and enrich
the open-source eco system. At it's worst you have organisations such
as Progress executing on predatory amoral tactics. The debate on
whether all corporations are ultimately operating on degrees of
amorality is out of scope for today :) As each year goes buy you start
to appreciate Red Hat more for the sterling work it does in open
source. While it would be nice if Progress got involved and
contributed to Drools, the reality is that this doesn't both me as
much as you think it would. It hasn't lestened the Drools community
and technology, we still go from strength to strength, and I've always
been more concerned about what we are doing and achieving than what
someone else may or may not be doing with regards to taking advantage
of our work.
Ultimately this is actually quite flattering and re-affirms the
strength of our technology and the directions we are going. I'd rather
have an organisation using Drools and moving it one step closing to
being a defacto standard, than they partner with someone else or
develop yet another rule engine splintering the market and confusing
users. "Drools Everywhere" has alwas been my moto :)
However that said, anyone who knows me, knows I have a mischievous
sense of humour and if Progress and Savvion aren't even going to at
the very least accredit us then they are fair game and surely myself
and the Drools community deserve a little fun at their expense :)
Mark - The Mischievous Lead
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
5009
days inactive
5009
days old
7 comments
4 participants
participants (4)
-
Edson Tirelli -
Greg Barton -
Mark Proctor -
Michael Neale