Preparing BMC Atrium CMDB for synchronization
At the time of its release, integration of the current version of BMC Atrium Discovery was supported with the following versions of BMC Atrium CMDB:
- 9.0 – Fully supported. There are some performance issues between BMC Atrium Discovery 10.2 and CMDB 9.0. These are resolved in CMDB 9.0 SP1 and CMDB 9.1.
If you intend to use CMDB sync, BMC recommends that you use BMC Atrium Discovery version 10.2.0.2 or later.
For more information, check the BMC Solution and Product Availability and Compatibility utility (SPAC) (support login required).
Before you begin
Before you can synchronize data to a supported BMC Atrium CMDB version, you must complete the following required tasks in the given order:
Integrating with earlier versions of BMC Atrium CMDB
BMC Atrium CMDB versions earlier than 7.6.04 are no longer supported and are not recommended for use. Synchronizing data to these earlier versions required you to complete some additional tasks. For more information, see the BMC Atrium Discovery 8.3 documentation.
- Create the BMC.ADDM dataset: The BMC.ADDM dataset must be manually created in the CMDB before a synchronization is attempted.
- Create the Job to merge the BMC.ADDM dataset with BMC.ASSET: After the BMC.ADDM dataset has been created, you must then create the job to reconcile it with the BMC.ASSET dataset.
- Check the BMC.ADDM dataset configuration: When using ITSM 7.0 with a backwards compatibility patch, you also need to ensure that the BMC.ADDM dataset is trusted.
After completing the preceding tasks, you can start synchronizing BMC Atrium Discovery data to BMC Atrium CMDB.
To obtain the maximum synchronization performance when using CMDB synchronization with BMC Atrium Discovery version 8.3.00 and later, you should consider tuning the database which BMC Atrium CMDB (or BMC Remedy AR System) is using.
Max-Entries-Per-Query parameter on BMC Remedy AR System server
To prevent BMC Discovery generating duplicate CIs in the BMC.ADDM dataset when running a resynchronization, you must set the BMC Remedy AR System server parameter
Max-Entries-Per-Query to at least 5000.
For more information, see the following documentation corresponding to your product version:
- BMC Remedy AR System 8.1 Performance tuning for Business Service Management
- BMC Remedy AR System 8.0 Performance tuning for Business Service Management
- BMC Remedy AR System Server 7.6 Performance Tuning for Business Service Management White Paper
The following sections in the White Paper are particularly relevant:
- Tuning an Oracle server.
- Best practices for tuning Oracle database servers.
- Tuning a SQL Server database.
- Best practices for tuning SQL Server database servers.
To create the BMC.ADDM dataset
Creating a BMC.ADDM dataset provides a dataset into which BMC Atrium Discovery is exported before being taken into the BMC Atrium CMDB. You create the dataset by using the BMC Atrium CMDB Reconciliation console. It is recommended that you create the BMC.ADDM dataset on the primary AR System server.
The dataset you are creating is intended to receive data from only one BMC Atrium Discovery appliance. If your deployment architecture requires that several scanners or consolidator synchronize to the same CMDB, then you must create a separate dataset for each. For example,
BMC.ADDM2, and associated reconciliation jobs.
Once created, do not make any changes to any CMDB synchronization dataset.
- You can create a dataset from one of two places in the console.
- From the Identify activity, click Add Dataset Identification Group Association,
and then click Create Dataset.
- From the Mid Tier, Open Atrium Core Console > Applications > Reconciliation console, and then click Create Dataset .
- From the Identify activity, click Add Dataset Identification Group Association,
Complete the following fields:
The name for the dataset. Usually set to
The system identifier for the dataset. This must match the ID used in the configuration within BMC Atrium Discovery. The default is
Set to "Writable".
Set to "Regular".
For more information, see:
To create the Identification, Merge and Purge Job
BMC Atrium Discovery exports its data to a staging dataset in BMC Atrium CMDB. The content of this dataset should then be reconciled into the BMC.ASSET dataset. To accomplish this, a new Job must be added in the BMC Atrium CMDB Reconciliation Console. It is recommended that you create the Identification, Merge and Purge Job on the primary AR System server.
To add the Identification, Merge and Purge Job
- From the Mid Tier, Open BMC Atrium Core Console > Applications > Reconciliation Application.
- Select the Create Standard Identification and Merge Job Icon
- Give the job a meaningful name (for example, "BMC ADDM - Identification, Merge and Purge").
- Set the Source Dataset to be "BMC.ADDM".
- After the job is Created, select BMC ADDM - Identification, Merge and Purge Job.
- Click the Edit Job icon.
- From the Job Editor window, in the Activities pane, select the New tab.
- In the New Activity pane, select Type as 'Purge', Sequence as '700' and Name as BMC ADDM - Purge.
- In the Datasets option, select Datasets as BMC.ADDM.
- Click Done.
- Click Save in the Job Editor.
The recommended practice is to create a precedence group that specifies a precedence for the BMC.ADDM dataset. The value depends on your relative preference for data populated by BMC Atrium Discovery and other data providers to the CMDB. For more information about how precedence groups are configured, see the following documentation corresponding to your product version:
- Defining a Precedence Association Set for reconciliation Merge activities for BMC Atrium Core 8.1
- Defining a Precedence Association Set for reconciliation Merge activities for BMC Atrium Core 8.0
- BMC Atrium CMDB 7.6.04 Normalization and Reconciliation Guide – section named "Defining a Precedence Association Set".
To check the BMC.ADDM dataset configuration
When using BMC Remedy ITSM 7.0 with a backward compatibility patch, you also need to ensure that the BMC.ADDM dataset is trusted. To do this:
- Access the Home page from BMC Remedy User or a browser.
- Access the Application Administration Console.
- Go to the Custom Configuration page.
- From Foundation, select Products / Operational Catalogs.
- Select Trusted Datasets for Product Catalog and click Open.
- Select the BMC.ADDM dataset for Dataset Name.
- Click Save.
To prepare synchronization to another CMDB
There might be occasions where you have already synchronized data to a CMDB, and you want to push data to a different CMDB. For example, if you are migrating from a development to a production environment for your CMDB, you might want to populate your new CMDB with fresh inventory data, independent of previous experiments you might have performed in the development instance. To synchronize the data to the new CMDB, perform the following steps using continuous CMDB synchronization.
- Stop any discovery runs.
- Stop continuous synchronization (wait for it to finish processing the queue).
- Change your configuration to the new CMDB dataset (see the tasks to complete before you begin).
- From the CMDB Sync > Configuration tab, click Test Connection to the new CMDB dataset.
Ensure that the connection is successful before proceeding.
- From the Infrastructure > Hosts page, send one or two hosts to the CMDB, and select Actions > Sync with CMDB.
- Start continuous synchronization.
- Restart your discovery runs.
The data is inserted into the new CMDB dataset as your Discovery jobs are executed. BMC Atrium Discovery detects that the data does not exist, and then inserts the data.
To synchronize data from multiple BMC providers to the CMDB
When multiple BMC providers synchronize data to the CMDB (such as BMC Atrium Discovery or BMC BladeLogic), there is the possibility that BMC_OperatingSystem instances might not be merged during reconciliation.
BMC Atrium Discovery is an agentless discovery tool and, as such, it might retrieve slightly different information than an agent-based product (for example, "SunOS" or "Solaris" for the same server). To eliminate this problem, you can add a fallback reconciliation rule based on the ClassId for BMC_OperatingSystem instances.