Annotating the SELECT keyword, the #NoCache annotation allows for deactivating the cache associated with this request.
This annotations allows to override the cache configuration as defined in the configuration file of the icCube server (i.e., icCube.xml within the installation bin directory).
Query
//#NoCache
WITH MEMBER [Measures].[foo] AS [Measures].[Amount]
SELECT [Measures].[foo] ON 0 FROM [Sales]
Result
Assertion : MDX Equals
WITH MEMBER [Measures].[foo] AS [Measures].[Amount]
SELECT [Measures].[foo] ON 0 FROM [Sales]
Annotating the SELECT keyword, the #Cache annotation allows for activating the cache associated with this request.
In this example, the maximum number of tuples being cached is defined as 42'000. There's no typical value for this parameter as this is highly dependent on the request. The icCube server log file is containing cache usage information (e.g., hit, miss, etc) for each request.
The parameter 'mode' defines what must be cached.
- ALL : is caching all the evaluated tuples.
- CALC_MEMBER_ONLY : is caching only calculated measures and members.
- ON_DEMAND : is caching measures and calculated members/measures whose REQUEST_CACHED cell property is defined.
The optional assertConsistency parameter allows for troubleshooting the request consistency. Bear in mind that activating this parameter means that the request will not benefit from the cache regarding its speed of processing. Indeed, every cached value is being re-computed.
This annotations allows to override the
cache configuration as defined in the configuration file of the icCube server (i.e., icCube.xml within the installation bin directory).
Query
//#Cache( mode = 'ALL', maxSize = 42000, assertConsistency = true )
WITH MEMBER [Measures].[foo] AS [Measures].[Amount]
SELECT [Measures].[foo] ON 0 FROM [Sales]
Result
Assertion : MDX Equals
WITH MEMBER [Measures].[foo] AS [Measures].[Amount]
SELECT [Measures].[foo] ON 0 FROM [Sales]