Restarting an apply request

If the operating environment or DBMS fails while the High-speed Apply Engine is running, you can use this procedure to restart the apply request from the point of failure.

Before you begin

To successfully restart an apply request, you must:

  • Know the name or alias of the restart table that tracked the original request
  • Know the restart ID that you specified or that High-speed Apply Engine generated for the job
  • Specify the appropriate restart parameters in the configuration, as described in Specifying-restart-parameters

Related topics

Enabling-restart-processing

Specifying-and-using-restart-parameters

Warning

High-speed Apply Engine bases its checkpoint and restart processing on the number and sequence of records in the input file. If you modify an input file between the initial run of an apply request and the subsequent restart and you change the number or sequence of records in the file, the results of the restarted run are unpredictable.

In general, do not modify the input file of an apply request between initial run and restart. Modify the configuration file only to the extent required to resolve error conditions. Ensure that changes to the configuration do not alter the distribution of work between agents. For example, do not change the values of configuration parameters such as MaxAgents, DistributionType, PartitionClustering, or RIClustering.

To restart an apply request

  1. Review the display output file or your job output to determine whether there is a restart ID that you can use for the apply request. The restart ID is either specified in the configuration or generated by High-speed Apply Engine (in message BMCAPT0097).

    Important

    The display output file is specified with the FileName (DFN) parameter in the [DisplayOutput] section.

    1. If you find a restart ID, skip to Step 2.
    2. If no restart ID exists for the request, the job cannot be restarted from the point of failure.You must analyze the target tables to determine where apply processing ended. Contact BMC Support to determine if you can use conflict parameters effectively to permit a restart of this apply request.
  2. Edit the configuration for the apply request.
    1. If you directed High-speed Apply Engine to generate a restart ID (no value was specified for the RestartID parameter), set the RestartType parameter to Restart and the RestartID parameter to the generated restart ID. (Obtain the generated value from the display output file or from message BMCAPT0097).
    2. If you specified a value for the RestartID parameter, verify that the RestartID parameter contains your specified value. Use one of the following techniques:

      • Verify that the RestartType parameter is New/Restart.

        The advantage of this technique is that you do not have to change your configuration when you restart the apply request. The disadvantage is that you can encounter negative consequences if you accidentally run an apply request in error. For example, if you attempt to apply a new input file but the previous run did not complete successfully, you can fail to apply part of the input file. If you attempt to restart the previous input file but the previous run did complete successfully, you can encounter -803 SQL codes, or corrupt the data in your database.

      • Set the RestartType parameter to Restart.

        The advantage of this technique is that you encounter fewer negative consequences when you run an apply request in error. The disadvantage is that you must change your configuration to restart an apply request.

  3. Verify that the restart table is available for the apply request.
  4. Submit the apply request to restart apply processing. For more information about using restart parameters, see RestartID or RestartType. 

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*