IRE Best Practices

The fact is there is no “one size fits all” approach or set of configurations that would be the best for everyone that we can share. What we can share are some key, general recommendations, specific examples, and how best to handle those examples that are found frequently in ServiceNow.

For more in-depth information on the CMDB Identification and Reconciliation you should refer to ServiceNow’s documentation.

Identification Rules: Identification rules should not be changed lightly. Carefully consider, before making a change to an identification rule, whether the issue should be resolved another way, such as data quality or governance/process.

Inclusion Rules: Use inclusion rules only if necessary. Typically, inclusion rules are only needed in the event you have another issue you’re trying to address such as duplication. For example, if you have a retired server and a new server that both share a name, and you want to ensure the old server is not updated you would look to inclusion rules to help solve that issue.

However, inclusion rules have a cost in terms of performance as they add another check that records must go through. You could also unintentionally cause other impacts such as making duplication more common if processes that update state are not followed correctly, for example.

Do use inclusion rules if you need to! If your environment has governance/processes that allow for name to be reused for retired CIs, then using inclusion rules will likely be necessary and reduce the chance of incorrectly updating the wrong CI. So, if you need them, use them.

Performance Impact Considerations: As mentioned before, inclusion rules add additional checks during the identification process. This can sometimes have a minor impact on how fast records process.

Reconciliation Rules: We recommend that you review the out-of-the-box reconciliation rules. In most cases they may be sufficient to handle your specific use case; however, if you have a complex CMDB with multiple sources of data populating the CMDB, you may need to add additional rules to avoid attribute flapping or duplication of data.

Source Identification: Clearly label the source of each CI so that you can track back issues and updates to the original source system. As an example, if you have ITOM discovery running, the source would be ServiceNow. If you also have the Flexera One connector populating the CMDB, the source would be Flexera One. This practice allows you to easily see which source has been updating the CI—which can be helpful for auditing and data quality.

Normalization Rules: Leverage IRE's normalization capabilities to ensure that the incoming data adheres to a standardized format—making it easier to manage and use.

Deduplication: Use the Identification Rules within IRE to ensure that incoming CIs are not duplicates of existing records. Configure these rules to take multiple sources into account.

Conditional Transform Maps: Make use of ServiceNow's robust transform maps to conditionally process incoming data. For example, you might want to transform incoming data differently based on its source or the type of CI.

Collaboration with Stakeholders: Involve teams responsible for data sources to establish alignment on rules and priorities.

Prioritization of Sources: In a multi-source environment, conflicts can arise when the same CI is discovered or managed by multiple tools. Use IRE to establish source-precedence rules.

10.

Batch Handling: When integrating large volumes of data, use batch processing to minimize performance impact. Set up IntegrationHUB to collect and push data in batches rather than in real-time, if acceptable.

11.

Reconciliation Policies: Implement reconciliation policies in IRE that determine how conflicts between data from different sources are resolved.

12.

Error Handling: Implement robust error-handling and logging mechanisms to capture and report errors in the integration flows. Ensure that these are monitored regularly for prompt resolution.

13.

Validation and Testing: Before pushing data from a new source into your CMDB, validate it in a test environment. Make sure the IntegrationHUB flows and transform maps are thoroughly tested with representative data sets. In the case of Flexera One connector you can set the integration property Enable Test Load in the Integration Properties page and set the volume of records to 30k. This setting will help speed up the process should you find any issues with the configuration of IRE and need to adjust.