The 4 key steps to implement a CMDB

TLDR: There are 4 key steps to setting up an effective CMDB: planning and design, data collection, configuration and customization, and finally continuous maintenance and management. These steps are essential to the success of the CMDB: a properly populated CMDB should enable better planning of the IT strategy and associated project management.

If you're not yet comfortable with the concept of CMDB, you can read this article What is a CMDB?

In this article, we will see how Mama’s Little Bakery (MLB), a leading European cheesecake producer, an SME of 500 employees, starts and sets up its CMDB.

Step 1: Planning and Design

Let's go! The first step to setting up a CMDB is to plan and design its structure. This means defining the objectives and requirements of the CMDB, while identifying the data sources that will feed the database. You also need to determine what types of configuration items (CIs) will be included in the CMDB, such as servers, software, applications, networks, and much more.

During this phase, it is important to involve key stakeholders, including IT teams, business managers, and end users, to ensure alignment between the CMDB and the needs of the entire organization. You will also need to establish governance processes to ensure the quality and integrity of CMDB data.

So who do we invite to the kickoff?

• The CIO, because they are very likely the sponsor of the operation. The process of building a CMDB serves them directly for the proper planning of their IT strategy.

• The Architect! (Well, if you have one...). We expect them to work the most on the subject. However, we shouldn't forget the following.

• Project Managers. Whether Business Manager or project manager for a CRM deployment, they are the knowledgeable ones on the application and business layer.

• The Ops. In the operations team, they master technical components like servers, databases, and they will also be able to make the link with the applications running on them.

At MLB, one of the most sensitive parts is Human Resources. By mutual agreement with management, the CIO decided to start by modeling the entire HRIS from server to applications to identify the most critical processes and applications. Information will be collected from IT technical tools (monitoring tools, incident management tools, VCenter...) but also by launching a collection from HR business managers (Recruitment, Payroll Management, Administrative Management). The IT department gives itself 3 months to complete this project.

Step 2: Data Collection

Once planning is complete, you can move on to data collection to feed the CMDB. This can involve integrating data from various sources, such as incident management tools, change management tools, network monitoring tools, automated configuration scans, etc.

It is important to ensure that collected data is accurate, complete, and up to date. To do this, you will need to clean and normalize the data before importing it into the CMDB. Documentation of data sources and integration processes is also essential to ensure data traceability and reliability.

We can rely for example on:

• An extraction from your monitoring platform (Nagios, PRTG, Zabbix): You will have a good proportion of servers and equipment that count in the IS (or else, they are not monitored...)

• An RVTools inventory to take a snapshot of the vCenter inventory. It's quite classic but it feels good to say it.

• A company organization chart: We've all seen one! And this is the time to capitalize on it to define the company's business services and especially the managers of each.

At MLB, the data collection phase is a crucial step. After defining the objectives and requirements of the CMDB, the IT department developed a Python script to collect and gather all data from previously identified sources into an Excel file. On the other hand, a Typeform form was sent to each HR business manager to collect the list of applications used to perform all their tasks. It is a hybrid IS: home-made and publisher applications hosted on a private cloud but also a set of SaaS applications.

Step 3: Configuration and Customization

Once data has been collected, you can proceed with configuring and customizing the CMDB. This involves defining configuration item (CI) classes and relationships between them, as well as configuring attributes and values for each CI. You will also need to define version and dependency management rules to ensure the accuracy of information in the CMDB.

During this phase, it is important to maintain an iterative approach, regularly testing and validating the CMDB configuration to ensure it meets the organization's needs. You may also need to train end users and CMDB administrators on its use and features.

Once data is collected, it is now a matter of modeling it in the CMDB. Form data has been compiled and script data cleaned and normalized. For simplification choice, MLB decided that only processes, applications, servers, and networks will be represented. Middleware like Enterprise Service Bus / ESB will not be represented in the CMDB. Throughout the data integration process, the manager regularly sends a restitution to business managers for validation.

Step 4: Maintenance and Continuous Management

Once the CMDB is operational, it is crucial to put in place maintenance and continuous management processes to ensure its quality and usefulness in the long term. This involves regularly monitoring and updating CMDB data, resolving errors and inconsistencies, and responding to changes in the organization's IT environment.

It is also important to regularly evaluate the effectiveness of the CMDB against the objectives defined during initial planning, and to make necessary adjustments to optimize its performance. Continuous collaboration with stakeholders and establishing a data governance culture are also essential to ensure the long-term success of the CMDB.

We have a review process at MLB, largely inspired by the ISO9001 standard. So in the control plan, we have planned to review the content of the CMDB monthly. When work is followed regularly, the update effort is not that great and we have constituted our different "views" on the IS to identify the latest modifications (VMs created less than 1 month ago, new applications in Azure SSO, etc.)

That’s it!

Setting up an effective CMDB requires careful planning, precise data collection, careful configuration, and continuous management. By following these four key steps, you can ensure your CMDB is well-designed and constitutes a valuable tool for your organization's IT management.