[Mondrian] Mondrian System Limitations

[Matt Campbell]

FWIW, we have had good experience with very large schemas:  1200+ shared
dimensions, 300+ base measures, 2500+ calculated members, 20-30 base cubes.
 Some of these dimensions also have high cardinality:  100s of thousands of
dimension members in some cases.  In production we typically allocate 20G
to Mondrian running inside of tomcat, but these schemas will run
efficiently with just 1G allocated in development environments.


On Tue, Nov 27, 2012 at 9:09 PM, Julian Hyde <jhyde at pentaho.com> wrote:

> There are no hard limits to the number of cubes, dimensions, hierarchies,
> members etc., other than available memory and maybe the size of a Java list
> or array (2 ^ 31). Obviously nothing is for free, so all else being equal,
> a large schema will be slower than a small one. But Mondrian is designed to
> degrade gracefully.
>
> Julian
>
>
> On Nov 27, 2012, at 10:42 AM, Taylor Curro <patriotsfan1236 at gmail.com>
> wrote:
>
> Good Afternoon Julian-
>
> I was wondering if you could answer a quick thought I have been trying to
> figure out.  I saw online a comparison of OLAP servers and some of the
> servers list out their system limits (number of cubes, measures,
> dimensions, hierarchies in dimension, levels in hierarchy, and dimension
> members) and could not seem to find any information regarding the Mondrian
> OLAP server.
>
> I have been focusing primarily on the system limits for the number of
> measures and dimensions (though I am trying to figure out all areas listed
> above), since cubes could be scaled-out since they do not interact with
> each other (correct me if I'm wrong).  I am trying to see how complex the
> data can be when it comes to dimensions and cardinality to prevent scale
> problems.  I was wondering if you could provide any guidance/insight into
> this topic?  I would greatly appreciate it.
>
> Thank you,
>
> -Taylor
> patriotsfan1236 at gmail.com
> _______________________________________________
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20121128/2ccb3478/attachment.html