Decision Management & Intelligence (ODM, DI)

Decision Management & Intelligence (ODM, DI)

Connect with experts and peers to elevate technical expertise, solve problems and share insights


#Data
#Data
#Businessautomation
 View Only
Expand all | Collapse all

ODM Decision Warehouse database is growing so large the DB server running out of disk space. Please advise.

  • 1.  ODM Decision Warehouse database is growing so large the DB server running out of disk space. Please advise.

    Posted Mon July 26, 2021 10:55 PM

    We have many rulesets within many rulepapps deployed to RES and execution traces enabled for many of the rulesets within the Decision Warehouse DB.

    The DB has grown so large in volume and continues to grow that the disks are at capacity on the DB server and we are seeing errors: "No Space Left on Device" . Since we no longer need ruleset execution tracing, what is the procedure to disable the ruleset execution traces and to reduce the size of the DW db in order to re-claim the storage that was used by DW ?



    #OperationalDecisionManager(ODM)
    #Support
    #SupportMigration


  • 2.  RE: ODM Decision Warehouse database is growing so large the DB server running out of disk space. Please advise.

    Posted Wed July 28, 2021 06:30 PM

    First thing to do is verify which rulesets have execution tracing enabled, which will fill up the data warehouse tablespace RESDWTS quickly.

    In RES console, Explore tab: for each ruleset within a deployed ruleapp,

    check if Enable tracing in Decision Warehouse , i.e. checkbox.

    Disable it by unchecking the check box and click save.

    Either repeat for each ruleset within each ruleapp. This will disable all DW execution server traces to stop the fast growth of the DB.

    Or if you are using the ruleset execution traces, then you will want to

    selectively disable certain monitoring events for specific ruleapps and also optimize the DB in order to prevent the RESDWTS tablespace from reaching storage capacity.

    Once you have disabled or limited the amount of new DW records from being adding to the DW table: EXECUTION_TRACES, it then needs to be cleaned or truncated.

    Note: For DB2 database, reorg/runstats will also need to run in order to reclaim the storage previously used by the tablespace.

    There are a few options to clean or truncate the DW table: EXECUTION_TRACES

    1) RES console: Navigate to Decision Warehouse tab, Clear Decisions Task

    URL: http://<host_or_IP><port>/res/protected/bam_cleanup.jsf

    Here you can manually clear the the decision traces.

    Refer to the ODM Doc for details:

    https://www.ibm.com/docs/en/odm/8.9.2?topic=help-monitoring-managing-decision-warehouse

    2) REST API: You can use the RES REST API to programmatically accomplish the same task as in 1 above by using the deleteDecisionTraces REST API call.

    https://www.ibm.com/docs/en/odm/8.9.2?topic=SSQP76_8.9.2/com.ibm.odm.dserver.rules.ref.res/html/restapi/html/decisiontraces_ecmapi.html

    You can access the REST API Test Tool in RES console, to test out the REST API calls.

    URL: http://<host_or_IP><port>;/res/protected/rest.jsf

    then click on /decisiontraces

    There are 2 REST calls you can use to delete the decision traces.

    /decisiontraces

    /decisiontraces/{execution_id}

    https://www.ibm.com/docs/en/odm/8.9.2?topic=tool-using-rest-api-test

    3) Another option is to truncate the DB2 table, EXECUTION_TRACES directly.

    Normally it is not recommended that customers modify the production DB tables/data directly, but in this case of DW it is safe to do so, since EXECUTION_TRACE table is the only one for the decision warehouse module and the records are not linked to the other tables in the decision server database.

    It is advised to have a full backup of the db prior to truncating the table.

    The db2 SQL statement for truncate would look something similar to this:

    db2> TRUNCATE TABLE EXECUTION_TRACES IGNORE DELETE TRIGGERS DROP STORAGE IMMEDIATE

    Note: You will need to perform a db2 REORG operation against the EXECUTION_TRACES table to reclaim the table spaces.

    This is because the table space is not automatically reclaimed in DB2 after you have delete the table rows.

    For details, please refer to: https://www.ibm.com/support/pages/how-reclaim-space-after-dropping-indexes-or-rows-db2-table

    The storage capacity should now be reclaimed to the OS and the "No Space Left on Device" errors should no longer be present in the logs.



    #OperationalDecisionManager(ODM)
    #Support
    #SupportMigration