![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
if the following performance issue still an issue in SSAS2005 Please read the 'limit the use of complex named sets' paragraph in the following link. http://www.microsoft.com/technet/pro.../ansvcspg.mspx -------------------------------- Limit the Use of Complex Named Sets Named sets are useful for sharing common definitions, such as the current month or last eight quarters. A named set can contain a constant set of members or an expression that resolves to a set. Named sets are evaluated and resolved at connection time for each client connection, regardless of whether the client actually uses the named set in a query. Experience has shown that simple named sets perform well, but that complex named sets (such as an expression filtering for the top five customers) can reduce query responsiveness, because each complex named set requires the scanning of many cells by Analysis Services at connection time. For example, if you have one million customers and have a named set containing an expression that evaluates the top five customers, Analysis Services must evaluate all one million customers to determine the top five customers for each client that establishes a connection. This query is a very expensive query against a large cube with many dimension members. If you have multiple named sets, each must be resolved at connection time before the client can begin querying. To improve performance in this scenario, consider enumerating the top five customers in the named set definition, rather than calculating the top five customers with every connection. You can then periodically update the contents of the named set by changing the named set definition (this does not require any processing). |
![]() |
| Thread Tools | |
| Display Modes | |
| |