[Mondrian] 300 sql queries on a 'simple' query , or , AxisMemberList.getTopParent behaving weird

Pedro Alves pmgalves at gmail.com
Thu Aug 23 08:04:10 EDT 2012


At least this served for something, now I know how to write test cases 
for my own stuff :p


http://pedroalves-bi.blogspot.de/2012/08/mondrian-writing-test-cases.html

On 08/23/2012 10:45 AM, Pedro Alves wrote:
>
> *uff*
>
>
> turns out that I actually had hasAll = true, but didn't issue a refresh
> solution rep. Only noticed that after setting up a test case in
> mondrian. ffs.... :X
>
>
>
>
> On Wed 22 Aug 2012 08:35:52 PM WEST, Paul Stoellberger wrote:
>> do you have an all member in your hierarchy? no idea if thats relevant
>> though
>>
>>
>>
>>
>> Am 22.08.2012 um 20:33 schrieb Pedro Alves <pmgalves at gmail.com>:
>>
>>>
>>>
>>> In pentaho master:
>>>
>>> The query:
>>>
>>> SELECT NON EMPTY [Browsers].[Major].Members ON COLUMNS,
>>> [Locations].[Location].Members ON ROWS FROM [Market Share]
>>>
>>>
>>> does 230 sql queries, one for each member of locations.location.
>>>
>>>
>>> SELECT NON EMPTY [Store].[Store Country].Members ON COLUMNS,
>>> [Product].[Product Name].Members ON ROWS FROM [Warehouse]
>>>
>>> on steelwheels does not.
>>>
>>>
>>> After a day of debugging, I traced the issue to
>>> AxisMemberList.getTopParent that's called from
>>> AxisMemberList.mergeMember.
>>>
>>>
>>> On foodmart, every [Product].[Product Name] member has [Product].[All
>>> Products] as the getTopParent.
>>>
>>>
>>> On mine, every location has itself as the top parent, resulting in an
>>> axisMembers (on RolapResults) with 230 elements. Foodmart example has
>>> only 2 ([Store].[All Stores],[Products].[All Products])
>>>
>>>
>>> What am I missing? What causes the getTopParent to return null?
>>>
>>>
>>>
>>> -pedro
>>> _______________________________________________
>>> Mondrian mailing list
>>> Mondrian at pentaho.org
>>> http://lists.pentaho.org/mailman/listinfo/mondrian
>> _______________________________________________
>> Mondrian mailing list
>> Mondrian at pentaho.org
>> http://lists.pentaho.org/mailman/listinfo/mondrian


More information about the Mondrian mailing list