This documentation supports the 9.1 to 9.1 Service Pack 3 version and its patches of BMC Atrium Core. The documentation for version 9.1.04 and its patches is available here.

To view the latest version, select the version from the Product version menu.

Best Practices for handling Reconciliation


To view the best practices video for Reconciliation, see Reconciliation Engine Best Practices . This video is recorded using the earlier version of BMC Atrium Core and is valid for BMC Atrium Core 9.1. 

Reconciliation identifies, merges, and compares CIs. Purpose of reconciliation is to take disparate data sources and merge them into a unified view. There are some important best practices to follow for the identify activity in Reconciliation. Before you start reconciling CIs ensure that you have performed the tasks listed below.

Before you begin

  • Ensure that you have completed the initial data load,
  • Ensure that the CIs are residing in the respective datasets.
  • Ensure that you have Normalized datasets.

Best Practices

  • Always perform the identify activity against the source datasets and never against the production datasets. Do not specify production as a dataset while performing the identify activity because the CIs in the production dataset are already identified. The production dataset is the result of the reconciliation activities. 

  • Do not include the production as a source dataset in the merge operation. You should merge sources to get the resultant production dataset. You should merge sources to get the target or the production dataset. 

  • Continuous reconciliation should be used only after the initial load of CIs is complete. After the initial bulk load is complete, if at a later stage you have a changes coming in at a pace at which you need to have the information regularly updated in Atrium CMDB, use the continuous reconciliation option. For example, in a virtual environment where you have frequent changes where you need the information in real time. 

  • Use the appropriate reconciliation mode based on the behavior of the data source.

    For example, in case of first time reconciliation of data source and larger data volumes (in ~100K to millions), use batch mode to process data. For data sources which provides continuous updates (ongoing changes) and smaller data volume (in few hundreds to thousands) use a continuous mode.

  • In Reconciliation Merge activity,  use merge order "By Class In separate Transaction" to get maximum throughput of the activity.
  • BMC Atrium CMDB experts recommend that you use the Defer if NULL feature, this powerful feature should be uses in conjunction with precedence rules. For a detailed conceptual information on this feature see, Handling NULL values.


Ensure parallel reconciliation jobs are not dependent on each other to prevent data integrity issues.

Related topics

Overview of reconciliation

Was this page helpful? Yes No Submitting... Thank you