What is the size of the largest dimension that MS OLAP supports

What is the size of the largest dimension that MS OLAP supports

Post by Simon Le » Thu, 12 Dec 2002 16:11:55



I am loading a cube and its shared dimensions currently. I have a file
which has about 4 million records for a dimension it is a WBS Element
dimensions appears Analysis Server is complaining that I have exceeded
the maximum number of dimensions.

How many dimensions records can I load max into a SQL Server
dimension.
Appears the select distinct count started and appeared it counted
until 98,000 or something of that sort and it abends stating maximum
number of dimensions exceeded. Any idea how to fix this I have quite a
few such large dimensions that I need to use.

Thanks
Simon

 
 
 

What is the size of the largest dimension that MS OLAP supports

Post by Chris Webb [MS » Thu, 12 Dec 2002 19:29:15


Hi Simon,

The theoretical limit for the number of members in a MOLAP dimension is
2147483647 (which I think is what you're talking about, not the maximum
number of dimensions in a cube, which is 128 - see the BOL topic 'Analysis
Services-specification'), but the practical upper limit is dictated by
memory. 4 million members in a dimension is possible, but I would advise the
following:
- Make sure you have 4Gb of memory on your server, and that you use the /3Gb
Windows boot setting and adjust the HighMemoryLimit registry setting to make
sure you can use all available memory
- Only include member properties that you need to use. Each dimension member
will use approx 125b *plus* the size of any member properties you have
defined, and on large dimensions you need to be careful not to waste memory
- Use the smallest data type possible whenever you can, eg for keys

You can see a more detailed giude to how dimension memory is used in this
post, from earlier in the year:
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&threadm=e...
SsBHA.2660%40tkmsftngp05&rnum=1&prev=/groups%3Fq%3DMosha%2Bdimension%2Bsize%
2Bestimate%2Bgroup:microsoft.public.sqlserver.olap%26hl%3Den%26lr%3D%26ie%3D
UTF-8%26oe%3DUTF-8%26selm%3DeETIR%2524SsBHA.2660%2540tkmsftngp05%26rnum%3D1

Also, are you sure you're not hitting the 64000 maximum number of children
for one member? If so, then you can get around it by creating member
groups - see the topic 'Member Groups' in BOL.

Finally, the obvious question to ask before you go to all the effort of
creating such a large dimension is - do your users really need to see the
whole of the dimension? For example, if your 4 million members are all
customers, is it likely that they need to view data at the customer level in
the cube? They may be happy with data at a higher level, and then maybe
using drillthrough to see the detail data.

HTH,

Chris

--

Chris Webb
Consultant, Microsoft Services Switzerland
(to email: remove 'online.' from the address this was posted with)

Microsoft Services Switzerland and our partners can help you with your
Analysis Services project!

Disclaimer: This posting is provided 'AS IS' with no warranties, and confers
no rights.


Quote:> I am loading a cube and its shared dimensions currently. I have a file
> which has about 4 million records for a dimension it is a WBS Element
> dimensions appears Analysis Server is complaining that I have exceeded
> the maximum number of dimensions.

> How many dimensions records can I load max into a SQL Server
> dimension.
> Appears the select distinct count started and appeared it counted
> until 98,000 or something of that sort and it abends stating maximum
> number of dimensions exceeded. Any idea how to fix this I have quite a
> few such large dimensions that I need to use.

> Thanks
> Simon


 
 
 

1. Cognos & MS OLAP Services - large dimension

We're using Powerplay Enterprise Server 6.6 & OLAP Services 7.0 sp2. One of
our dimensions has 113,000 leaf rows, but performs particularly badly & uses
masses of memory on the web server when i drill through the hierarchy. It
even times out when i try to move 'down a level' to the leaf nodes. The
situation is markedly better when i use it in a fairly small cube, but it
still fails when i try to move 'down a level'.

Has anybody tried using larger dimensions than this & what sort of issues
did they face? Any tips on how i can overcome the problems & what i should
try to avoid?

Thanks in advance,

Mike Weaver

2. QuickTab to File?

3. Help please ! MS SQL -> Oracle !

4. FoxBase+/Mac: yet _another_ problem...

5. Large Dimensions kill OLAP performance.

6. US-DE: Wilmington-Strong Oracle 8i/9i Programmer with XML

7. Table Size and OLAP dimension

8. Q: BrioQuery, MS OLAP Cubes: Dimension/Measure Access Control

9. Time dimension not sorted in MS Olap

10. Manual Level in Dimension using MS OLAP

11. Big dimensions / MS OLAP