Showing posts with label groups. Show all posts
Showing posts with label groups. Show all posts

Friday, March 30, 2012

One large database or many small....

Hi Experts

We are debating what is best:

1. To combine all the company's data in one large database, and use schemas and file groups to create logical and physical distribution on drives and namespaces

or

2. Distribute the data into smaller databases with related data - eg. products and product description in one db, Customers in another and orders and orderlines in a third db.

Just what are the pros and cons?

regards

Jens Chr

In almost any case I would suggest one database. The ability to use foreign key constraints is enough right there. The only reason I would suggest many databases is if you might need to seperate them into different servers, but that is very seldom.

It is just so much easier to work with data in the same database versus multiple databases. You get one log, which is easier to manage (unless you have tons of disk drive channels to have multiple log channels, it will be faster to have the one log). If you seperate stuff into filegroups instead of databases, you can still backup and restore them independently if you need to, as well as position the tables in a filegroup on different drives as needed.

|||I echo Louis' sentiments - listen to him. He's a wise and experienced SQL person :)

Check out my SQL Server 2005 Video Tutorials: http://www.learnsqlserver.com/

Friday, March 9, 2012

OLE DB timeout processing AS Cube

When doing a full process of the Analysis Services Cube with several partitioned measure groups. We receive: "OLE DB error: OLE DB or ODBC error: Query timeout expired HYT00" or "OLE DB error: OLE DB or ODBC error: Query timeout expired HY008". There are 410 million rows in one base table and 80 million rows in another base table. The other measure groups process fine. It has abended at 61 minutes the last 3 times. I can individually process each measure group successfully.

Any help is greatly appreciated, Edwina Derrick

Simular problem for me. Have a huge database to be queried and can't pass the 60 minutes query time despite having made sure that ALL!! timeout values are suposed to be eternal (zero in most cases). This seems like a bug to me. I used the developer release version of the SQL2005. Would appriciate the Microsoft development team to test this problem and to suply information of when and how to bypass this 1 hour timeout limitation.

|||Try increasing ExternalCommandTimeout Analysis Server advanced property.

Being in situation when Analysis Server needs to wait for an hour for some data to come from SQL Server is very unusual. Such situation ususally means that SQL server needs to perform some heavy operation before it can start sending data out. Try optimizing your SQL Server database.
Edward Melomed (MSFT)


This posting is provided "AS IS" with no warranties, and confers no rights.|||

WTF....

isn't there a way to set the timeout?

If I was able to "Optimize my SQL Server database".....

I have five fact tables (All of them are required) and each table has it's own Dimensions.

since I am unable to create a cube on a left join relationship I need to work with views ("SQL server needs to perform some heavy operation before it can start sending data out").

OLE DB timeout processing AS Cube

When doing a full process of the Analysis Services Cube with several partitioned measure groups. We receive: "OLE DB error: OLE DB or ODBC error: Query timeout expired HYT00" or "OLE DB error: OLE DB or ODBC error: Query timeout expired HY008". There are 410 million rows in one base table and 80 million rows in another base table. The other measure groups process fine. It has abended at 61 minutes the last 3 times. I can individually process each measure group successfully.

Any help is greatly appreciated, Edwina Derrick

Simular problem for me. Have a huge database to be queried and can't pass the 60 minutes query time despite having made sure that ALL!! timeout values are suposed to be eternal (zero in most cases). This seems like a bug to me. I used the developer release version of the SQL2005. Would appriciate the Microsoft development team to test this problem and to suply information of when and how to bypass this 1 hour timeout limitation.

|||Try increasing ExternalCommandTimeout Analysis Server advanced property.

Being in situation when Analysis Server needs to wait for an hour for some data to come from SQL Server is very unusual. Such situation ususally means that SQL server needs to perform some heavy operation before it can start sending data out. Try optimizing your SQL Server database.
Edward Melomed (MSFT)


This posting is provided "AS IS" with no warranties, and confers no rights.|||

WTF....

isn't there a way to set the timeout?

If I was able to "Optimize my SQL Server database".....

I have five fact tables (All of them are required) and each table has it's own Dimensions.

since I am unable to create a cube on a left join relationship I need to work with views ("SQL server needs to perform some heavy operation before it can start sending data out").

OLE DB timeout processing AS Cube

When doing a full process of the Analysis Services Cube with several partitioned measure groups. We receive: "OLE DB error: OLE DB or ODBC error: Query timeout expired HYT00" or "OLE DB error: OLE DB or ODBC error: Query timeout expired HY008". There are 410 million rows in one base table and 80 million rows in another base table. The other measure groups process fine. It has abended at 61 minutes the last 3 times. I can individually process each measure group successfully.

Any help is greatly appreciated, Edwina Derrick

Simular problem for me. Have a huge database to be queried and can't pass the 60 minutes query time despite having made sure that ALL!! timeout values are suposed to be eternal (zero in most cases). This seems like a bug to me. I used the developer release version of the SQL2005. Would appriciate the Microsoft development team to test this problem and to suply information of when and how to bypass this 1 hour timeout limitation.

|||Try increasing ExternalCommandTimeout Analysis Server advanced property.

Being in situation when Analysis Server needs to wait for an hour for some data to come from SQL Server is very unusual. Such situation ususally means that SQL server needs to perform some heavy operation before it can start sending data out. Try optimizing your SQL Server database.
Edward Melomed (MSFT)


This posting is provided "AS IS" with no warranties, and confers no rights.|||

WTF....

isn't there a way to set the timeout?

If I was able to "Optimize my SQL Server database".....

I have five fact tables (All of them are required) and each table has it's own Dimensions.

since I am unable to create a cube on a left join relationship I need to work with views ("SQL server needs to perform some heavy operation before it can start sending data out").