|
|
|
date: Thu, 3 Jul 2008 00:30:36 -0700 (PDT),
group: microsoft.public.sqlserver.olap
back
Re: To many Cubes and dimensions?
Well one minute was to process one cube not all 150. If I place the
cube in an different database , but on the same server and the same
disks it will process in a few seconds. So it seems that having many
cubes in the same database slows down the processing time. It looks
like it takes 1 minute to actually start processing when I have 150
cubes in the same database.
On Jul 5, 3:55 am, "Jeje" wrote:
> well... the server has limited resources...
> first, what are your expectations? what is your target? (in seconds)
>
> you can spread the cube among different disks (insure that the raid
> controller is setup for 100% write cache), add CPU, use real time cubes
> instead of MOLAP cubes, improve the number of parallel processing tasks
>
> but 1 minute is nothing for 150 cubes.
>
> wrote in message
>
> news:0829a0f3-b235-4048-a5cc-00df16113acf@y38g2000hsy.googlegroups.com...
>
>
>
> > I have an OLAP database with 150 cubes and each of the cubes have 12
> > dimensions. Response times of MDX queries are fine, but Processing
> > times of the cubes seems to be increasing as I add more cubes. If the
> > cube is the only one in a database it takes only a few seconds to
> > process, but when its in a database with lots of other cubes it takes
> > about 1 minute to process (This is before any data is actually in the
> > fact table). Anyone have any ideas on how to solw this problem?
>
> > Magne- Hide quoted text -
>
> - Show quoted text -
date: Wed, 9 Jul 2008 05:23:57 -0700 (PDT)
author: Magne
Re: To many Cubes and dimensions?
when SSAS process a cube, a transaction is made
larger the database is , slower the transaction is.
try to play with the advanced processing options to use multiple
transactions instead of 1
"Magne" wrote in message
news:e6e04773-ceb0-4338-89a8-fdd198e1e707@x35g2000hsb.googlegroups.com...
> Well one minute was to process one cube not all 150. If I place the
> cube in an different database , but on the same server and the same
> disks it will process in a few seconds. So it seems that having many
> cubes in the same database slows down the processing time. It looks
> like it takes 1 minute to actually start processing when I have 150
> cubes in the same database.
>
> On Jul 5, 3:55 am, "Jeje" wrote:
>> well... the server has limited resources...
>> first, what are your expectations? what is your target? (in seconds)
>>
>> you can spread the cube among different disks (insure that the raid
>> controller is setup for 100% write cache), add CPU, use real time cubes
>> instead of MOLAP cubes, improve the number of parallel processing tasks
>>
>> but 1 minute is nothing for 150 cubes.
>>
>> wrote in message
>>
>> news:0829a0f3-b235-4048-a5cc-00df16113acf@y38g2000hsy.googlegroups.com...
>>
>>
>>
>> > I have an OLAP database with 150 cubes and each of the cubes have 12
>> > dimensions. Response times of MDX queries are fine, but Processing
>> > times of the cubes seems to be increasing as I add more cubes. If the
>> > cube is the only one in a database it takes only a few seconds to
>> > process, but when its in a database with lots of other cubes it takes
>> > about 1 minute to process (This is before any data is actually in the
>> > fact table). Anyone have any ideas on how to solw this problem?
>>
>> > Magne- Hide quoted text -
>>
>> - Show quoted text -
>
date: Thu, 10 Jul 2008 00:02:30 -0400
author: Jeje
|
|
