PDA

View Full Version : [Mondrian] bug [ 1651936 ] NonEmptyCrossJoin Stack Overflow



Will Gorman
03-29-2007, 03:40 PM
Hello, this message is primarily to Richard but anyone's help would be
appreciated.

I'm working on the bug [ 1651936 ] stack overflow for slicer on
NonEmptyCrossJoin named set.

The infinite loop seems to occur inside
CrossJoinFunDef.nonEmptyListNEW(). There is call to
CrossJoinFunDef.checkData(), which ends up calling
evaluator.evaluateCurrent(). evaluteCurrent() ends up making it's way
back into nonEmptyListNEW().

I'm guessing we'll need to update checkData() in some way, but I'm not
quite sure what exactly needs to be changed.

Thanks for your help!

Will

http://sourceforge.net/tracker/index.php?func=detail&aid=1651936&group_id=35302&atid=414613

_______________________________________________
Mondrian mailing list
Mondrian (AT) pentaho (DOT) org
http://lists.pentaho.org/mailman/listinfo/mondrian

Richard Emberson
03-29-2007, 03:51 PM
Get the current Mondrian p4 head and see if this happens.

Richard


Will Gorman wrote:
> Hello, this message is primarily to Richard but anyone's help would be
> appreciated.
>
> I'm working on the bug [ 1651936 ] stack overflow for slicer on
> NonEmptyCrossJoin named set.
>
> The infinite loop seems to occur inside
> CrossJoinFunDef.nonEmptyListNEW(). There is call to
> CrossJoinFunDef.checkData(), which ends up calling
> evaluator.evaluateCurrent(). evaluteCurrent() ends up making it's way
> back into nonEmptyListNEW().
>
> I'm guessing we'll need to update checkData() in some way, but I'm not
> quite sure what exactly needs to be changed.
>
> Thanks for your help!
>
> Will
>
> http://sourceforge.net/tracker/index.php?func=detail&aid=1651936&group_id=35302&atid=414613
>
> _______________________________________________
> Mondrian mailing list
> Mondrian (AT) pentaho (DOT) org
> http://lists.pentaho.org/mailman/listinfo/mondrian
>


--
Quis custodiet ipsos custodes:
This email message is for the sole use of the intended recipient(s) and
may contain confidential information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all
copies of the original message.
_______________________________________________
Mondrian mailing list
Mondrian (AT) pentaho (DOT) org
http://lists.pentaho.org/mailman/listinfo/mondrian

Richard Emberson
03-29-2007, 04:20 PM
Are you in a position to work on the source code? or
can you create an example of the error with the
FoodMart dataset?

Richard


Richard Emberson wrote:
> Get the current Mondrian p4 head and see if this happens.
>
> Richard
>
>
> Will Gorman wrote:
>> Hello, this message is primarily to Richard but anyone's help would be
>> appreciated.
>>
>> I'm working on the bug [ 1651936 ] stack overflow for slicer on
>> NonEmptyCrossJoin named set.
>> The infinite loop seems to occur inside
>> CrossJoinFunDef.nonEmptyListNEW(). There is call to
>> CrossJoinFunDef.checkData(), which ends up calling
>> evaluator.evaluateCurrent(). evaluteCurrent() ends up making it's way
>> back into nonEmptyListNEW().
>>
>> I'm guessing we'll need to update checkData() in some way, but I'm not
>> quite sure what exactly needs to be changed.
>>
>> Thanks for your help!
>>
>> Will
>>
>> http://sourceforge.net/tracker/index.php?func=detail&aid=1651936&group_id=35302&atid=414613
>>
>>
>> _______________________________________________
>> Mondrian mailing list
>> Mondrian (AT) pentaho (DOT) org
>> http://lists.pentaho.org/mailman/listinfo/mondrian
>>
>
>


--
Quis custodiet ipsos custodes:
This email message is for the sole use of the intended recipient(s) and
may contain confidential information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all
copies of the original message.
_______________________________________________
Mondrian mailing list
Mondrian (AT) pentaho (DOT) org
http://lists.pentaho.org/mailman/listinfo/mondrian

Will Gorman
03-29-2007, 04:30 PM
There is a test case already checked in:

NonEmptyTest.testVCNativeCJWithIsEmptyOnMeasure()

You'll need to set a flag in the mondrian.properties file:

mondrian.native.crossjoin.enable=false

I'm in a position to work on the source.

Will

On Thu, 2007-03-29 at 13:22 -0700, Richard Emberson wrote:
> Are you in a position to work on the source code? or
> can you create an example of the error with the
> FoodMart dataset?
>
> Richard
>
>
> Richard Emberson wrote:
> > Get the current Mondrian p4 head and see if this happens.
> >
> > Richard
> >
> >
> > Will Gorman wrote:
> >> Hello, this message is primarily to Richard but anyone's help would be
> >> appreciated.
> >>
> >> I'm working on the bug [ 1651936 ] stack overflow for slicer on
> >> NonEmptyCrossJoin named set.
> >> The infinite loop seems to occur inside
> >> CrossJoinFunDef.nonEmptyListNEW(). There is call to
> >> CrossJoinFunDef.checkData(), which ends up calling
> >> evaluator.evaluateCurrent(). evaluteCurrent() ends up making it's way
> >> back into nonEmptyListNEW().
> >>
> >> I'm guessing we'll need to update checkData() in some way, but I'm not
> >> quite sure what exactly needs to be changed.
> >>
> >> Thanks for your help!
> >>
> >> Will
> >>
> >> http://sourceforge.net/tracker/index.php?func=detail&aid=1651936&group_id=35302&atid=414613
> >>
> >>
> >> _______________________________________________
> >> Mondrian mailing list
> >> Mondrian (AT) pentaho (DOT) org
> >> http://lists.pentaho.org/mailman/listinfo/mondrian
> >>
> >
> >
>
>

_______________________________________________
Mondrian mailing list
Mondrian (AT) pentaho (DOT) org
http://lists.pentaho.org/mailman/listinfo/mondrian

Richard Emberson
03-29-2007, 05:10 PM
The two quick things that I thought might give me some insight
failed to do so (I tried them setting the property to false).
So, I don't off hand know what could be wrong.

Richard


Will Gorman wrote:
> There is a test case already checked in:
>
> NonEmptyTest.testVCNativeCJWithIsEmptyOnMeasure()
>
> You'll need to set a flag in the mondrian.properties file:
>
> mondrian.native.crossjoin.enable=false
>
> I'm in a position to work on the source.
>
> Will
>
> On Thu, 2007-03-29 at 13:22 -0700, Richard Emberson wrote:
>> Are you in a position to work on the source code? or
>> can you create an example of the error with the
>> FoodMart dataset?
>>
>> Richard
>>
>>
>> Richard Emberson wrote:
>>> Get the current Mondrian p4 head and see if this happens.
>>>
>>> Richard
>>>
>>>
>>> Will Gorman wrote:
>>>> Hello, this message is primarily to Richard but anyone's help would be
>>>> appreciated.
>>>>
>>>> I'm working on the bug [ 1651936 ] stack overflow for slicer on
>>>> NonEmptyCrossJoin named set.
>>>> The infinite loop seems to occur inside
>>>> CrossJoinFunDef.nonEmptyListNEW(). There is call to
>>>> CrossJoinFunDef.checkData(), which ends up calling
>>>> evaluator.evaluateCurrent(). evaluteCurrent() ends up making it's way
>>>> back into nonEmptyListNEW().
>>>>
>>>> I'm guessing we'll need to update checkData() in some way, but I'm not
>>>> quite sure what exactly needs to be changed.
>>>>
>>>> Thanks for your help!
>>>>
>>>> Will
>>>>
>>>> http://sourceforge.net/tracker/index.php?func=detail&aid=1651936&group_id=35302&atid=414613
>>>>
>>>>
>>>> _______________________________________________
>>>> Mondrian mailing list
>>>> Mondrian (AT) pentaho (DOT) org
>>>> http://lists.pentaho.org/mailman/listinfo/mondrian
>>>>
>>>
>>
>
> _______________________________________________
> Mondrian mailing list
> Mondrian (AT) pentaho (DOT) org
> http://lists.pentaho.org/mailman/listinfo/mondrian
>


--
Quis custodiet ipsos custodes:
This email message is for the sole use of the intended recipient(s) and
may contain confidential information. Any unauthorized review, use,
disclosure or distribution is prohibited. If you are not the intended
recipient, please contact the sender by reply email and destroy all
copies of the original message.
_______________________________________________
Mondrian mailing list
Mondrian (AT) pentaho (DOT) org
http://lists.pentaho.org/mailman/listinfo/mondrian