Hi Edward,
What is wrong with my cube?
The dbo.Prescription table is in the data source view.
Please advice
Ronald
You are receiving a message during processing of one of your partitions: 'Prescrition'
The error indicates your server cannot find the source table for your partition.
In the previous post I suggested you modify source table for partition 'Prescrition'.
Have you tried it? Did it work?
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
The strange thing is that- i create two cubes with same data source using the same fact table and the same dimension in both cases.
However with the first one i dont define the dimension as a time dimension and the cube processes fine.
With the second cube i do define the dimension as a time dimension and then the cube processing errors. The error say attribute key cannnot be found? But it was found in the first cube?
Anyone know what might be causing this?
Thanks,
Can you try and use copy/paste functionality in BI Dev studio to make a copy of your existing cube and try to process it then?
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
Thanks for the reply unfortunatly i still get he same error when i try this.
Some more information-
The error i get is that attribute key with value 0 cannot be found. i get as many of these errors as there are nulls in the fact tables date dimension key column. There arent any 0 values.
So it might be that for some reaon the nulls are converted to zeros when a date dimension is added and this throws the referential integrity errors? Please help..
Thanks,
|||I think you guessed it right.
You need to make sure your fact table doesnt have null's in the date dimension key column.
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
But surely my date dimension field can contain nulls. It is a nullable field in the OLTP system.?
i noticed this http://mgarner.wordpress.com/tag/mdx/ see the 'null = 0' post at the bottom.
|||This is not a good practice to have a nullable key.
Although Analysis Services allows you to map null keys to Unknown member in the dimension, it is not a good idea.
To solve this problem try define a named calculation in your DSV to map null keys to some pre-defined value on the fly.
Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.
Thanks for you reply Edward
But.
Analysis services is not converting nulls to unknown if the column is defined as a time dimension. It is converting them to zeros and raisng referential integrity errors.
Is it the case that Analysis services cubes cannot contain nullable time dimensions?
This doesnt sound right to me as date columns often signify an event and therefore a null indicates that the event hasnt occured.
Not being able to define nullable time dimensions if true is a serious flaw in analysis services 2005.
What value could i set the calculated field to in the case of nulls. Surely unknown is the right value for a null date? I certainly wouldnt want to put in someone arbitary date and presumaably it would have to be a vlaue of datetime data type?
[Edit]
Edward below is some test code which builds a test fact table and test time dim. If you define the dim as a time dim then the processing will error if you insert the null in the time dim. Note if you dont define it as a time dim it processess fine with or without the null.
Please have a look at this and let me know.
Thanks alot for all your help.
if object_id('DateDim') is not null
drop table DateDim
go
if object_id('datefact') is not null
drop table datefact
go
create table DateDim
(ID int identity, date smalldatetime)
go
insert DateDim
select getdate()
union all
select getdate()+1
union all
select getdate() +2
union all
select getdate() +3
union all
select getdate() +4
union all
select getdate() +5
create table datefact
(id int identity, datekey int , amount int)
go
insert datefact
select 1, 50
union all
select 1, 50
union all
select 3, 50
union all
select 4, 50
union all
select 5, 50
union all
select 5, 50
/*--Uncomment this so a null is insert into the time dimension reprocess the cube and it will error
union all
select null, 50
*/
sql
We are in the process of migrating to a new 64-bit server. We changed the data provider for the cube from "Microsoft OLE DB Provider for Oracle" to ".Net Provider\OracleClient Data Provider" for the 64-bit and when we try and process this cube, we get the following error:
Errors in the high-level relational engine. The following exception occurred while the managed IDataReader interface was being used: Value was either too large or too small for a Currency..
This very same cube was able to process on the 32-bit server just fine. The field that the error refers to is a number(28,0) in oracle and is fully populated in the database. I set up a simple cube on my laptop and if I cut the value in the field down to 15 numbers it will process correctly. I have maxed out the precision of the number type to 38 as well and I am still getting the error when more than 15 numbers are populated in a single field.
Does anyone know what is keeping the cube from processing with ODP.Net when it is processing fine with the other data provider?
This has to do with the conversion of that data type in the Oracle Provider, Analysis Services wouldn't be generating that error, it would be just passing it on. I would consult with Oracle to find out more regarding it.
Is the data type in the dimension or Fact for this field Currency? If so you may want to change it in AS so there is no conversion.
David
Hi guys! Could you help me with the next problem:
Failed to build the OLAP cubes. Error: Analysis Services session failed with the following error: Query (1, 10) Parser: The syntax for ',' is incorrect.
I have installed sql2005 sp2, but didn't get much
Pls. help
If you have a new issue, please create a new thread rather than just replying to an old one. You are more likely to get an answer.
It sounds like there might possibly be an extra comma in your MDX script. It looks like it is at character 10 on the first line (that's what the 1,10 means)
Hi,
I tried to create a cube file in Excel 2003. Data source is SSAS 2005.
On the last step of wizard I get the following error:
Microsoft OLE DB Provider for Analysis Services 2005:
Query (3,2) Parser: The syntax for 'DIMENSION' is incorrect.
Seems like a bug. Is there any fix?
Tried to search google, found nothing...
SQL Server 2005 Developer SP2
Excell 2003 SP2
Thx for any tip...
certainly sounds like a bug. Can you provide a repro' that the product team can look at?
thanks,
Bala.
Hi,
I tried to create a cube file in Excel 2003. Data source is SSAS 2005.
On the last step of wizard I get the following error:
Microsoft OLE DB Provider for Analysis Services 2005:
Query (3,2) Parser: The syntax for 'DIMENSION' is incorrect.
Seems like a bug. Is there any fix?
Tried to search google, found nothing...
SQL Server 2005 Developer SP2
Excell 2003 SP2
Thx for any tip...
certainly sounds like a bug. Can you provide a repro' that the product team can look at?
thanks,
Bala.
Hello,
I created an account called sa2 to obtain access to create a datasource for the AdventureworksDW sample database on a remote server, and granted this user all rights on the database.
When I was ready to deploy the cube I typed in the machine name (sserver) in connection settings, and then deployed. However, I got an error:
"The connection either timed out or was lost."
I then tried:
sserver\sserver
but got:
"Error 1 The project could not be deployed to the 'sserver\sserver' server because of the following connectivity problems : A connection cannot be made to redirector. Ensure that 'SQL Browser' service is running. To verify or update the name of the target server, right-click on the project in Solution Explorer, select Project Properties, click on the Deployment tab, and then enter the name of the server. 0 0
Permissions issue for the sa2 account?
Hello,
Yes, from Bids. Not sure at what point the error occurs, I select deploy, processing occurs and then the error message is displayed.
Since this post I dropped the firewall on the server, and the database seems to be ceated but errors occur, so deployment is not successful.
This link:
http://www.sqljunkies.com/WebLog/edwardm/archive/2006/05/26/21447.aspx
is suggesting that I can only login with a Windows account.
|||You need a Windows Account with administrator access to the Analysis Server to deploy the database (not necessarily an Windows Administrator account). You can use a SQL Server login for the connection from Analysis Services to the relational engine.|||Make sure you used the same windows login with remote server, same username and same password. I think the problem was in authentication method.Best regards,
Hery|||
Hello,
Thanks for the suggestions everyone. The computer from which I am trying to deploy does not login to the network, so I don't have a profile which I can apply to the database. As a workaround for now I will deploy to localhost.
