Tune social analytics indexing

Overview

Use SearchCellConfig to set the number of iterations for a job.

The work for these jobs is divided up based on iterations. To improve performance, we can configure the number of iterations specified for a particular job based on the needs of the deployment. For example, reducing the number of iterations results in faster performance but is more memory-intensive.

Tune the social analytics indexing process

1. Check out search-config.xml...
   cd app_server_root/profiles/Dmgr01/bin
   wsadmin.sh -lang jython
   execfile("searchAdmin.py")
   SearchCellConfig.checkOutConfig("/tmp", "cellName")

   To get cellName...

   print AdminControl.getCell()

2. Set the number of iterations used by a specified social analytics job...

   SearchCellConfig.setSandIndexerTuning(String indexer, Int iterations)

   ...where...

   indexer	Name of the social analytics indexing job. Valid values: evidence, graph, manageremployees, and tags.
   iterations	An integer that specifies the number of iterations for the specified social analytics indexing job.

   For example:

   SearchCellConfig.setSandIndexerTuning("manageremployees",200)
   SearchCellConfig.setSandIndexerTuning("graph",400)

3. Check in the updated search-config.xml file
   SearchCellConfig.checkInConfig()

4. Type exit.

5. Stop the server or servers hosting the Search application, and then restart the Search servers. The next time the scheduled task runs, persisted seedlists are retained after indexing finishes.

Parent topic:
Administer the social analytics service

Related:
Add scheduled tasks for the social analytics service