Is the CMDB Dead? 3 Ways to Guarantee a Successful Implementation

Posted by on October 25, 2016


Jarod GreeneOne of the biggest challenges IT organizations have faced over the last decade has been establishing a Configuration Management Database (CMDB). This certainly hasn’t been from lack of effort, as many authors, analysts, consultants, and vendors can easily put their children’s children through college on the strategies and solutions sold in pursuit of a single-source-of-truth from an IT infrastructure perspective. These efforts have not been frivolous. The value proposition of the CMDB has been well articulated. So why do so many IT organizations struggle with their CMDB initiatives? Before we dive into this question (and deliver some much-needed answers!), let’s make sure we are operating with the same definitions and context.

Tweet this: Why so many IT organizations struggle with their CMDB initiatives and how to achieve a smooth implementation

Definition and Purpose of the CMDB

At its core, the CMDB is a database that contains all relevant information about IT assets (referred to as Configuration Items, or CIs) installed across the infrastructure environment. The CMDB also contains data that describes the relationships and inter-connectivity between CIs.

The ability for IT organizations to understand how applications, servers, middleware, databases, networking devices, and storage devices interact to deliver business services is vital in anticipating the impact of infrastructure changes. As more organizations have come to realize 80 percent of service downtime is related to people or process (not errors in the infrastructure) and that almost all aspects of their business were becoming IT driven, the CMDB has offered a better way to manage change.

But, a funny thing happened on the way to the near-mythical 99.999% uptime: IT organizations found it difficult to trust the single-source-of-truth, for a variety of reasons.

Why CMDBs Often Fail to Deliver

In the 10 years I spent as an ITSM analyst at Gartner, I spoke to only about 20 organizations (out of many hundreds) who had success with their CMDB initiatives. I recently attended a session on this very topic at the Gartner IT Operations Solutions and Strategies Summit in National Harbor, Maryland. Gartner’s new lead analyst on CMDB, Hank Marquis, presented a very compelling presentation titled, “Renovating Your CMDB to Improve Business Outcomes.”

The session began by highlighting the fact that almost 80 percent of CMDB initiatives fail, and that on average organizations will fail three times over before a successful initiative. Marquis also quickly shared Gartner’s Strategic Planning Assumption which states that by 2019, failure rates will increase from 80 percent to 90 percent.

But why?

For starters, initial attempts to populate the CMDB often rely on inventory and discovery tools that simply aren’t up to the task. Poor reconciliation and normalization capabilities mean that redundant CI data getsinto the database, and without a consistent taxonomy, multiple items are reported in multiple ways. In other words, “garbage in, garbage out.”

Also, not much thought is given to synchronization—that is, making sure that changes made in the production environment are reflected in the CMDB. When change records aren’t matched against actual changes, it results in drift from the current state, rendering the CMDB untrustworthy, especially as the number of changes increase in frequency.

How to Revamp Your CMDB

With those odds, why do a CMDB project at all? In Hank Marquis’ presentation, he went on to outline the fact that as IT organizations become more complex—more mobile, more virtual, and more software-defined—the ability to understand your environment isn’t a nice to have, it’s an organizational imperative, particularly for digital business.

But despite the low success rates of first-time CMDB initiatives, the silver lining is this: when implemented effectively, the CMDB delivers the business value it promises—and more. While it isn’t easy, IT organizations that have been successful with their CMDBs reap the benefits of improved change success rates, less critical application downtime, and even the ability to demonstrate regulatory compliance with their CMDB.

So how do you position yourself for success with your CMDB efforts? Marquis provided simple and pragmatic advice for organizations wishing to revamp their CMDB.

Tweet this: 3 ways to position yourself for success with your CMDB efforts

1. Tie your CMDB goals to improved business outcomes

What sounds like a no-brainer is something many organizations fail to do: start with the business goal in mind, and work backwards to determine how the CMDB can affect that outcome.

Why this advice is key: Out of the box, the CMDB is an enabler with no intrinsic business value. The onus is on the IT organization to give it value, and that value must be tied to making the business better. A simple example is improving sales productivity by ensuring increased change success of the sales order processing application. This is more than feasible – it’s the role IT plays in supporting the business.

2. Be selective: don’t put everything in the CMDB

Gartner advises that customers only put as much IT asset information into the CMDB as is necessary to solve a business problem. On average, Gartner believes this to be ten to fifteen percent of the total assets, a stark contrast from IT organizations that want to discover everything and put it into the CMDB.

Why this advice is key: It speaks to the identification of the business goals from the beginning, whereby an IT organization might only need to model a few services to improve. Starting with one or two services could be a huge win for IT organizations, as opposed to dealing with the administrative challenges of a robust database which includes a lot of assets, such as endpoints, that don’t drive or deliver a service.

3. Use an Application Dependency Mapping Tool

To overcome the challenges associated with using point or domain discovery tools, the output of which need to be reconciled and normalized, Marquis suggests that IT organizations automate discovery and dependency mapping to shorten the time to value. Application Dependency Mapping software (also referred to as Discovery and Dependency Mapping [DDM] solutions) both discover the configuration item in the infrastructure, and map CI inter-connectivity, saving time and resources, while providing a more accurate view of the environment. In providing a partial list of vendors offering these capabilities, Marquis named FireScope as a viable solution for IT organizations in search of such a solution.

Why this advice is key: Discovery and Dependency Mapping solutions can aid organizations in maintaining an accurate and up-to-date view of the dependencies, so as to understand service relationships. The alternative manual approach can be highly labor intensive with time delays for updates and, again, run the risk of drift between the perceived state of the infrastructure in the CMDB and the actual state of the infrastructure in the production environment.

I wholeheartedly agree with Gartner’s recommendations, and would encourage any Gartner clients or prospects to engage with Marquis on this topic. CMDB initiatives can be a long, winding road with no end date in sight but, when done well and correctly, can provide significant value for IT organizations that wish to improve service performance and enable business outcomes. People and process play a significant role in these initiatives, as well as well-defined goals, the right technology, and the discipline to stay focused on only those aspects of the CMDB that will drive real business value.

Next Up: Cherwell Configuration Management Database

Establish a single source of truth across your IT infrastructure

Learn More

Learn More About Our Updates to the Cherwell Knowledge Articles