The Reconciliation Engine is a component installed with the BMC Atrium CMDB. The main function of the Reconciliation Engine is the creation of a production dataset that contains accurate data from all available sources. Data in the production dataset is then used by consuming applications.

For more information, see the following resources:

Before you run a Reconciliation Engine (RE) job, perform the following:

• Perform CDM metadata diagnostics - Run the CDMChecker utility to:
  • Detect invalid customization
  • Detect CDM corruption
    • For information about running this utility, see Using the CDMChecker tool.
    • For information about troubleshooting the errors, see Issues that might arise during the CDMChecker pre-upgrade check.
• Perform data diagnostics 
  
  • Run the cmdbdiag utility to check if data conforms with the model defined in BMC Atrium Configuration Management Database (BMC Atrium CMDB) and correct it. For information about using this utility, see Verifying your data model using the cmdbdiag program.
• Perform Reconciliation Engine-specific data diagnostics
  
  • Run the Data Analyzer tool to check whether the data follows the RE-specific guidelines. BMC highly recommends you to run this tool before running any RE job. See Using CMDB Data Analyzer to Investigate CMDB Data Issues.
• Perform manual checks for RE jobs
  
  • Make sure that all the classes and attributes used in standard or custom rules of all the RE jobs exist in your CDM data model. If a certain job is imported in your CDM, the job might fail.
  • Make sure that the data sets (source, target and precedence data set) referred in all the RE jobs also exist in the BMC_Dataset class.
• Check whether indexes are set
  
  • During upgrade, for optimal RE performance, ensure that the out-of-the-box indexes set up on BMC.CORE:BMC_BaseElement and BMC.Core:BMC_BaseRelationship forms.
  • Discrepancies, if any, are logged in the Atrium Installer log file. The following image displays an example message:
    

Watch the video on Reconciliation Engine best practices:

https://youtu.be/8dykPDe2GDU

Also, review the Best Practices for handling Reconciliation.

Configure recommended settings

Ensure that the System-level settings and Job-level settings are configured as per the recommendations. See the following images for recommended settings:

System-level settings

Job-level settings

For more information, see Configuring Reconciliation Engine system parameters.

 You should consider changing the default configuration settings in the following scenarios:

• If you are dealing with large volume of data, increase the value of the Definition Check Interval field.
• If you need specific log details, ensure that you change the log level to Debug. (Change the default log value only when you need and reset it back to default log level when you are done). Additionally, set the Maximum Job Log File Size value to 50000 KB.
• If the RE Job is not responding for a long period, set Job Idle Time value to the desired time other than the default value (0).

Standard identification and precedence rules simplify the process of creating reconciliation jobs. Standard rules use defaults for Identify and Merge activities and automate the creation of reconciliation jobs.

The standard rules work with all classes in the Common Data Model (CDM) and BMC extensions. They identify each class using attributes that typically have unique values, and they merge based on precedence set for BMC datasets. Standard reconciliation jobs always use these standard rules, but you can also use standard rules when you create your own reconciliation job.

You must use standard rules:

• When configuration items are published into the CMDB or when the data is pulled from the CMDB.
• To uniquely identify attributes of the discovered configuration items.

For more information, see the following topics:

• Determining when to use standard reconciliation jobs and rules
• Standard identification and precedence rules for reconciliation jobs
• Configuring standard identification rules for reconciliation
• Configuring standard precedence rules for reconciliation

A set of rules that are created as per your specific business requirements and are not part of the standard rules are termed as custom rules. You can use custom rules for CI classes.

The following is an example scenario in which you might want to create custom rules:

If you have extended the Common Data Model by adding a new attribute, which is specific to your requirement and which might not be populated by any discovery tool. You might want to define custom rules to identify your assets based on this attribute.

For more information about using custom rules, see Creating an identification ruleset for reconciliation.

Standard identification and precedence rules simplify the process of creating reconciliation jobs. A reconciliation job is provided with a standard rule sets of identification and precedence rules.

These standard rules work with all classes in the Common Data Model (CDM) and BMC extensions. They identify each class using attributes that typically have unique values, and they merge based on precedences set for BMC datasets. Standard reconciliation jobs always use these standard rules, but you can also use them when you create a custom reconciliation job.

For more information, see Standard identification and precedence rules for reconciliation jobs.

 Watch the following video:

For more information, see Configuring standard precedence rules for reconciliation.

• Do not set a RE ID manually in any dataset.
• Do not manually delete data, which is marked as soft delete.
• Do not delete datasets manually. You must also need to clean up dataset references from Reconciliation Engine, Normalization Engine etc

For most reconciliation activities, you can specify a qualification set for the purpose of restricting the instances that participate in an activity. Qualification sets, which are reusable between activities, are qualification rules that each select certain attribute values. Any instance that matches at least one qualification in a set can participate in an activity that specifies the qualification set.

For example, you might create a qualification set that selects instances that were discovered within the last 24 hours and have the domain "Frankfurt" if your company just opened a Frankfurt office and you are reconciling its discovered CIs for the first time.

• For more information, see Using Qualification Sets in reconciliation activities.
• For information about conventions to be used in qualification of Reconciliation Engine activity, see Qualification conventions for reconciliation activities.

The CMDB Data Analyzer is a collection of tools that enable you to perform data analysis and also identify data inconsistencies in any CMDB dataset.

For more information, see Using CMDB Data Analyzer to Investigate CMDB Data Issues.

As a best practice, BMC recommends that you must reconcile only those CIs that have been normalized or those that do not require normalization. To reconcile the appropriate CIs, enable the Process Normalized CIs Only option in the Job Editor.

To reconcile manually created CIs, perform the following steps:

1. Create a custom job.
2. Define the job for the classes in which you populate the data.
3. Identify the CIs.
   
   1. Ensure that you adhere to the identification rules for the discovered jobs so that your CI matches with the discovered CIs.
   2. Ensure that you populate at least those columns/attributes correctly that are defined in the identification rules (This helps in correct identification of the CI with the existing CI)
4. Set a lower precedence value for the manual dataset, but set a highest precedence to the manually edited columns/attributes of the dataset.

CIs are partially merged when the merge process fails due to data issues.

See the following flowchart that explains the checks to be performed if the CIs are partially merged:

 Run the CMDB Data Analyzer utility to find duplicate CIs. See Using CMDB Data Analyzer to Investigate CMDB Data Issues.

Run Reconciliation Engine's Identification activity across multiple datasets. For example, open the BMC.CORE:BaseElement form and enter the RE ID value and search. If a CI having the same RE ID is found in multiple datasets, it means that the same CI exists in those datasets.

Run the CMDB Data Analyzer utility to find orphan CIs. See Using CMDB Data Analyzer to Investigate CMDB Data Issues.

The following table lists the common RE issues and provides access to relevant resources and workaround to these issues.

To migrate the RE jobs, you must:

1. Export the RE job data using the Reconciliation Console. See Exporting reconciliation definitions.
2. Import the ARX data using BMC Remedy Data Import tool. See  Enabling the Data Import Utility in the BMC Remedy AR System online documentation.

The Reconciliation Engine process terminates because:

• The AR System server or database is busy and is unable to process the requests coming from RE.
• The AR System server is either not responding or not reachable.

In the above scenarios, RE awaits for a response from the AR System server and then terminates itself.

For more information, see the Knowledge Article.

Perform the following:

• Check whether the job schedule is deleted or is modified.
• Check whether the Application Pending form has an entry for that job.
  • If there is no entry for that job, enable the AR API, SQL and Filter logs.
  • If the entry exists, check the AR Dispatcher log.
• Check whether RE is registered with the AR Dispatcher.
• Check the number of entries in the Application Pending form.
  • If there are many entries in the form and if you do not need them, delete those entries from the form.
• Check whether the RE job entry is in the Queued state in the RE job run. If yes, delete the entry from the Queued state.
• If you are using RE in a server group environment, ensure that the Reconciliation-Engine-Suspended flag is set to False (F). Note: You cannot set this value manually.

Perform the following:

1. Enable server-side logs.

   1. Edit the log4j.properties file located in the <AR_Installation_Home>\midtier\WEB-INF\classes folder. Add the following lines:

      and
   2. (BMC Atrium Core 7.6.04 or earlier): Edit the <AR_HOME>\ARSystem\midtier\WEB-INF\flex\services-config.xml file.
      
      (BMC Atrium Core 8.0 or later): Edit the <AR_HOME>\ARSystem\midtier\WEB-INF\AtriumWidget\flex\services-config.xml file.
      
      In the services-config.xml file, modify the logging level from Warn to Debug.
      
      For example, change the following line:
      <target class="flex.messaging.log.ConsoleTarget" level="WARN">
      To
      <target class="flex.messaging.log.ConsoleTarget" level=“DEBUG">
2. Enable client-side logs.
   For information about enabling client-side logs, see BMC Atrium Core Console client-side logging.

 See this knowledge article.