[Mondrian] Discussion about drill through

Julian Hyde jhyde at pentaho.com
Thu Jun 6 15:36:06 EDT 2013


Seems like we're in consensus.

The general goal seems to be to return at the lowest level available -- and this we're a ROLAP engine, that is the fact level. A possible extension to the DRILLTHROUGH command (for both Mondrian and Microsoft) might be some indication of the desired rollup level. Not that I'm proposing that we do this right now.

Julian


On Jun 6, 2013, at 11:26 AM, Matt Campbell <mcampbell at pentaho.com<mailto:mcampbell at pentaho.com>> wrote:

Regarding aggregation:  I’ve been doing some research into how Analysis Services 2005+ handles drillthrough.  SSAS returns drillthrough results at the lowest granularity of the cube, not necessarily the detail fact rows.  I believe they did this in part because they wanted drillthrough in MOLAP mode to not require SQL queries, and also because it avoided potential security issues.  Microsoft has some design tips they offer to get drillthrough at a lower grain—e.g. define an alternate measure group at a lower granularity and use it as the target for drillthrough.  By default, though, results are rolled up at the granularity of the attributes in the measure group being queried.

That said, it seems like most people would expect drillthrough to bring back the lowest level of detail available, which for Mondrian would be unaggregated fact rows.



From: mondrian-bounces at pentaho.org<mailto:mondrian-bounces at pentaho.org> [mailto:mondrian-bounces at pentaho.org<mailto:bounces at pentaho.org>] On Behalf Of Andrew Petrov
Sent: Thursday, June 06, 2013 10:08 AM
To: mondrian at pentaho.org<mailto:mondrian at pentaho.org>
Subject: Re: [Mondrian] Discussion about drill through

I think we should follow this spec http://technet.microsoft.com/en-us/library/ms145964.aspx, so

1) If not specified, returned columns should include all granularity attributes for all dimensions related to the measure group of the specified measure
2) See spec
3) I think drillthrough and aggregation are opposite processes and we should aggregate nothing in drillthrough
_______________________________________________
Mondrian mailing list
Mondrian at pentaho.org<mailto:Mondrian at pentaho.org>
http://lists.pentaho.org/mailman/listinfo/mondrian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20130606/0538e882/attachment-0001.html 


More information about the Mondrian mailing list