Page Comparison

The indexing operation of the bulk of data handled maybe using almost all the available memory based on the taxanomy cache size specified in the <DAS_HOME>/repository/conf/analytics/analytics-config.xml file.

e.g., If the available memory is 3.6 GB and 2.9 GB of that is used by the indexing operation as shown in the sample given above.
 

To overcome this out of memory issue, increase the taxonomy cache size by following the procedure given below.

1. Install the latest WUM update.

2. Add the following configuration under analytics-dataservice-configuration in the <DAS_HOME>/repository/conf/analytics/analytics-config.xml file (if it does not already exist).

   Info
   When ths configuration is not added in the analytics-config.xml file, the default values shown below apply.

   Code Block
   language xml
   <taxonomy-writer-cache> <cacheType>DEFAULT</cacheType> <!--<LRUType>STRING</LRUType>--> <!--<cacheSize>4096</cacheSize>--> </taxonomy-writer-cache>

3. To save the large taxanomy indexes in the file system of DAS instead of in the memory, change the value of the cacheType parameter to LRU as shown below.
   <cacheType>LRU</cacheType>

   Info
   When the value specified for the cacheType parameter is DEFAULT, taxanomy indexes are saved in the DAS memory.

4. Uncomment the LRUType parameter.
   

   Info
   The LRU type can be STRING or HASHED. The STRING value ensures that indexing is correctly done without collision. The HASHED value does not ensure the same, but it uses less RAM.

    

5. Uncomment the cacheSize parameter. Then increase the value for this parameter as required.
6. Restart the server.