IBM FileNet P8, Version 5.2.1            

Modifying settings for the Content Platform Engine Bulk Import tool in a clustered environment

To run in high availability active/active mode, you configure and run the bulk import tool on each node of the cluster and point each instance to a shared working directory where EOB (batch) files are processed.

Procedure

  1. Modify the cebi.cfg file for each instance of bulk import.
    1. Modify the WorkingDirectory entry in the cebi.cfg file of each node to include the path to the shared working directory where EOB files are located.

      The user that starts bulk import must have read/write access to that location. The path can be absolute or relative and can be a local mount point, a local path, or a UNC path.

      The following examples for Windows show a UNC path and a local mount point:
      WorkingDirectory=\\example.com\EOB
      WorkingDirectory=Z:\EOB
    2. Create a unique journal file for each node or bulk import tool instance.

      By default the journals directory is created in the home directory that is specified with -h in the start script. Uncomment and modify the JournalDirectory entry in the cebi.cfg file for each node to point to a unique directory where the journals are created.

      ; Location of journal log file. 'journals' is added to the folder path and created if needed. 
      JournalDirectory=C:\local_location\
    3. As a tuning option, you can uncomment and modify the SleepInterval entry in the cebi.cfg file for each instance of bulk import and specify a low number. When you modify the SleepInterval setting, you set the amount of time that one node sleeps after the node checks for work in the shared working directory. If one bulk import instance looks for work when another instance locks the working directory, bulk import waits for the specified amount of time before it looks for work again. If set for a long time, batch processing is delayed. If the interval is increased, batch processing might be delayed. If the interval is decreased, more system resources are used.

      For example, set the SleepInterval for 20 seconds (the default setting is 60 seconds).

      ; Determines the time interval to rescan the working directory for EOB files 
      SleepInterval=20
  2. To avoid errors that can be caused by reprocessing of any in-process batches when you start bulk import tool in a clustered environment, you must specify -r in the start command.

    The following code is an example is for Windows:

    "C:\Program Files (x86)\IBM\FileNet\ContentEngine\tools\_jvm17\jre\bin\java" 
    -Dlog4j.configuration=file:log4j.properties -classpath" 
    "C:\Program Files (x86)\IBM\FileNet\ContentEngine\lib\Jace.jar;C:\Program Files 
    (x86)\IBM\FileNet\ContentEngine\lib\log4j.jar;C:\Program Files 
    (x86)\IBM\FileNet\ContentEngine\tools\CEBI\bulkimport.jar" bulkImport.BI_Start -r -h C:\CB521\config1
    Bulk import locks the working directory, selects, and locks up to 20 EOB files and then releases the lock on the working directory. Other bulk import instances can then lock and select EOB files to process. The first instance continues to process its selected batch files after the directory lock is released.


Last updated: October 2015
p8pit029.htm

© Copyright IBM Corporation 2015.