9:14 a.m.
Hi,
we have an application with a lot of rules in a cached rulebase. When we
profile our application we see that about 30% of the time is used in
rulebase.newStatlessSession(). The time is basically used loading classes.
The question is, is it possible to create a pool of StalessSessions and
reuse them, or will this create problems.
Regards
Espen
--
View this message in context:
http://drools-java-rules-engine.46999.n3.nabble.com/Reuse-statless-sessio...
Sent from the Drools - User mailing list archive at Nabble.com.
9:50 a.m.
Hi,
According to what I've read so far, you can reuse Stateless sessions.
In particlar I've seen unit tests creating a Stateless session in a
@BeforeClass method, then this session was re-used across multiple tests
methods "pushing" different data-sets into that unique session instance
using commands (created with the org.drools.command.CommandFactory class).
Regards,
Bruno.
espen a écrit :
Hi,
we have an application with a lot of rules in a cached rulebase. When we
profile our application we see that about 30% of the time is used in
rulebase.newStatlessSession(). The time is basically used loading classes.
The question is, is it possible to create a pool of StalessSessions and
reuse them, or will this create problems.
Regards
Espen
12:16 p.m.
I might be putting too simplistic an interpretation on this, but this is
what the Javadoc says for StatelessKnowledgeSession
"StatelessKnowledgeSession provides a convenience API, wrapping
StatefulKnowledgeSession. It avoids the need to call dispose(). Stateless
sessions do not support iterative insertions and fireAllRules from Java
code, the act of calling execute(...) is a single shot method that will
internally instantiate a StatefulKnowledgeSession, add all the user data and
execute user commands, call fireAllRules, and then call dispose()."
So I'm not sure why you just wouldn't just directly use a
StatefulKnowledgeSession instead of trying to re-use a stateless one. You
can then update the Facts via their handles and update or retract and
reinsert as needed.
Cheers,
Pat
On Tue, Oct 26, 2010 at 3:50 PM, Bruno Freudensprung <
bruno.freudensprung(a)temis.com> wrote:
Hi,
According to what I've read so far, you can reuse Stateless sessions.
In particlar I've seen unit tests creating a Stateless session in a
@BeforeClass method, then this session was re-used across multiple tests
methods "pushing" different data-sets into that unique session instance
using commands (created with the org.drools.command.CommandFactory class).
Regards,
Bruno.
espen a écrit :
> Hi,
>
> we have an application with a lot of rules in a cached rulebase. When we
> profile our application we see that about 30% of the time is used in
> rulebase.newStatlessSession(). The time is basically used loading
classes.
>
> The question is, is it possible to create a pool of StalessSessions and
> reuse them, or will this create problems.
>
> Regards
> Espen
>
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
2:47 a.m.
Alright... The test case was delivered with the following documentation
(extract of the "Drools JBoss Rules 5.0 Developer's Guide" book):
"org.drools.runtime.StatelessKnowledgeSession
This is a type of knowledge session that doesn't keep any state between
invocations (an invocation is a call of the execute method). From the
rules perspective, a stateless session is no different than a stateful
session---the rules look exactly the same. The benefit that
statelessness brings is that the rule engine can do more optimizations.
StatelessKnowledgeSession as well as stateful session support a command
interface (command design pattern). There is no need to dispose a
stateless session after it has been used."
Bruno.
Patrick van Kann a écrit :
I might be putting too simplistic an interpretation on this, but this
is what the Javadoc says for StatelessKnowledgeSession
"StatelessKnowledgeSession provides a convenience API, wrapping
StatefulKnowledgeSession. It avoids the need to call dispose().
Stateless sessions do not support iterative insertions and
fireAllRules from Java code, the act of calling execute(...) is a
single shot method that will internally instantiate a
StatefulKnowledgeSession, add all the user data and execute user
commands, call fireAllRules, and then call dispose()."
So I'm not sure why you just wouldn't just directly use a
StatefulKnowledgeSession instead of trying to re-use a stateless one.
You can then update the Facts via their handles and update or retract
and reinsert as needed.
Cheers,
Pat
On Tue, Oct 26, 2010 at 3:50 PM, Bruno Freudensprung
<bruno.freudensprung(a)temis.com <mailto:bruno.freudensprung@temis.com>>
wrote:
Hi,
According to what I've read so far, you can reuse Stateless sessions.
In particlar I've seen unit tests creating a Stateless session in a
@BeforeClass method, then this session was re-used across multiple
tests
methods "pushing" different data-sets into that unique session
instance
using commands (created with the org.drools.command.CommandFactory
class).
Regards,
Bruno.
espen a écrit :
> Hi,
>
> we have an application with a lot of rules in a cached rulebase.
When we
> profile our application we see that about 30% of the time is used in
> rulebase.newStatlessSession(). The time is basically used
loading classes.
>
> The question is, is it possible to create a pool of
StalessSessions and
> reuse them, or will this create problems.
>
> Regards
> Espen
>
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org <mailto:rules-users@lists.jboss.org>
https://lists.jboss.org/mailman/listinfo/rules-users
------------------------------------------------------------------------
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
5:59 a.m.
2010/10/27 Bruno Freudensprung <bruno.freudensprung(a)temis.com>:
Alright... The test case was delivered with the following documentation
(extract of the "Drools JBoss Rules 5.0 Developer's Guide" book):
"org.drools.runtime.StatelessKnowledgeSession
[snip]
The benefit that statelessness brings is that
the rule engine can do more optimizations.
Ha!
I've been trying to find out about exactly this reputed distinction between
stateful and stateless sessions, NOT to be confused with "sequential mode",
which does introduce optimizations.
No success so far.
-W
StatelessKnowledgeSession as well
as stateful session support a command interface (command design pattern).
There is no need to dispose a stateless session after it has been used."
Bruno.
8 a.m.
You are right; so far (still haven't finished reading the book) it is
unclear whether the stateless session really implements some kind of
optimization.
Bruno.
Wolfgang Laun a écrit :
Ha!
I've been trying to find out about exactly this reputed distinction between
stateful and stateless sessions, NOT to be confused with "sequential mode",
which does introduce optimizations.
No success so far.
-W
10:11 a.m.
I might be wrong, but I believe the optimization he's talking about is the
sequential mode.
Coming back to the original problem in this thread, I'm amazed that you have
30% of your time linked to the newStatelessSession for a cached
KnowledgeBase! Loading of classes as the explanation for this also sounds
weird... do you spend a lot of time with not using any session so that the
classes get unloaded?
As a rule of thumb, I would say, cache knowledge bases, but throw away
sessions. Session creation is meant to be fast.
Leo.
On Wed, Oct 27, 2010 at 3:00 PM, Bruno Freudensprung <
bruno.freudensprung(a)temis.com> wrote:
You are right; so far (still haven't finished reading the book) it is
unclear whether the stateless session really implements some kind of
optimization.
Bruno.
Wolfgang Laun a écrit :
> Ha!
> I've been trying to find out about exactly this reputed distinction
between
> stateful and stateless sessions, NOT to be confused with "sequential
mode",
> which does introduce optimizations.
>
> No success so far.
>
> -W
>
>
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
10:12 a.m.
*As a rule of thumb, I would say, cache knowledge bases, but throw away
sessions. Session creation is meant to be fast.
I'm talking about stateless session here :)
On Wed, Oct 27, 2010 at 5:11 PM, Leonardo Gomes
<leonardo.f.gomes(a)gmail.com>wrote:
I might be wrong, but I believe the optimization he's talking
about is the
sequential mode.
Coming back to the original problem in this thread, I'm amazed that you
have 30% of your time linked to the newStatelessSession for a cached
KnowledgeBase! Loading of classes as the explanation for this also sounds
weird... do you spend a lot of time with not using any session so that the
classes get unloaded?
As a rule of thumb, I would say, cache knowledge bases, but throw away
sessions. Session creation is meant to be fast.
Leo.
On Wed, Oct 27, 2010 at 3:00 PM, Bruno Freudensprung <
bruno.freudensprung(a)temis.com> wrote:
>
> You are right; so far (still haven't finished reading the book) it is
> unclear whether the stateless session really implements some kind of
> optimization.
>
> Bruno.
>
> Wolfgang Laun a écrit :
> > Ha!
> > I've been trying to find out about exactly this reputed distinction
> between
> > stateful and stateless sessions, NOT to be confused with "sequential
> mode",
> > which does introduce optimizations.
> >
> > No success so far.
> >
> > -W
> >
> >
>
> _______________________________________________
> rules-users mailing list
> rules-users(a)lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/rules-users
>
5141
days inactive
5142
days old
7 comments
5 participants
participants (5)
-
Bruno Freudensprung -
espen -
Leonardo Gomes -
Patrick van Kann -
Wolfgang Laun