CMDB is pivotal to driving various business outcomes across the org. For CMDB Admins and owners to facilitate this, it is critical that they be able to actively monitor the health of their CMDB data, and swiftly act on remediating any CMDB health issues. CMDB Health Dashboard is an out-of-box feature which was created to help CMDB admins and users do this easily and seamlessly.

Below are some guidelines and best practices which can help users get the most value out of this dashboard.

1. Determine which data needs to be monitored actively: CMDB can often house a lot of nitty-gritty details about your infrastructure, especially when the data is populated by sources such as Discovery. However, users can get the most value out of this dashboard by narrowing down the most critical data set that should be actively monitored via the health dashboard. This can be done by setting up inclusion rules in classes.
2. Act on failures: The goal of this dashboard is to surface health issues of CMDB, so that CMDB owners and admins can take actions on these and remediate them.
3. Failure thresholds: There is an out-of-box limit on the number of failures encountered per metric, at which point the evaluation halts. So, for instance, if “Required Attributes” metric for Completeness is getting evaluated and 100K failures are encountered for this metric, the evaluation of this metric halts at that point. If the system is running into these thresholds, then it is likely that this is indicative of a systemic issue which should be looked it rather than continuing to churn through the entire dataset.

FAQs

1. Why isn’t there a single aggregated score representing CMDB health?

Completeness, Compliance and Correctness are completely different pivots across which the health of CMDB is gauged. The tool allows users to evaluate these for different slices of CMDB data as well. Given how disparate the metrics are in what they measure and how different the underlying dataset that is being evaluated for them can be, aggregating these into a single data point can make it incomprehensible as to what is really going on under the covers and understand the actual state of the system.

2. What isn’t there a trend line for the scores?

Without any context available for a historical data point (how many CIs were evaluated, were there changes to inclusion rules, was a new data source onboarded) it is not possible to draw inference from the trendline. For this reason, the trendlines were done away with starting in Xanadu.

3. We don’t see “Top 10 X generator” chart in Xanadu. Why is that?

Over the years, CMDB has gone through an evolution within ServiceNow. The amount of data that gets stored has increased. The way ownership of CIs is defined is also more methodical. Given the above, having a global “Top X CIs” list does not provide a lot of value to the users from a health monitoring perspective. Additionally, these were expensive indicators from a performance standpoint, not delivering enough value to the users. Hence, these were removed when the dashboards were rebuilt for Xanadu release.