Contact ITInvolve
x 


CMDB Inaccuracy

Finally a CMDB your team can trust.
Problems_CMDB-Inaccuracy_COLOR
Your organization has made a significant investment in implementing a configuration management database (CMDB). It was supposed to provide a single source of truth about your IT environment and make changes easier to plan with fewer unintended consequences. It was supposed to help you resolve issues faster, make compliance reporting easier, and provide a lot of other really important benefits.

But hardly anyone is using it and the most common reasons team members give is that it’s not up-to-date, it’s not accurate, or it’s incomplete. (Or perhaps you are still considering implementing a ITSM_Best-in-Class_grayscaleCMDB for the first time, and you’ve heard about these types of struggles from your peers and want to know how to avoid them).

In Our Experience

Most companies have tried to confront these issues by using auto-discovery tools to regularly update the CMDB, force every change through a formal change process that updates the CMDB, or create a CMDB “cleansing team” whose job it is to maintain the CMDB. These approaches help, but they fall short of delivering a living, trusted CMDB.

  • Discovery tools can only discover certain things off the wire

Finding the assets on the network isn’t so hard, but understanding the relationships between what’s running on them is much harder. Discovery tools generally have to be programmed to recognize specific patterns so that when a pattern is found, it can be recorded. This approach falls down when you have custom applications or your discovery tool has a limited set of patterns (none will cover 100% of what you have running).

  • It’s just not realistic to have every change go through a formal process and be approved by a committee

That’s a surefire way to slow an IT organization down and upset the business. Plus, just because you made the update in the CMDB doesn’t mean it’s been executed or was executed exactly as intended. Bottom line, if your CMDB isn’t accurate or trusted, then you’re getting a false sense of security when planning changes using this method too.

  • CMDB “cleansing teams” aren’t made up of your experts, because the experts are needed elsewhere

The cleansing team has less knowledge about the specifics of your environment so your experts do not trust their work. They do their best, but often consume a lot of expert cycles interviewing or emailing for information and then trying to update the CMDB with this information based on their limited understanding.

Discovery Tools and Traditional CMDBs Aren’t Enough

There’s a lot of very valuable dependency and relationship information that you simply can’t discover off the wire or store in a traditional CMDB.

For example, discovery tools can’t tell you which policies govern a given application and what time the back up runs so you can avoid an upgrade during that window. Nor can a CMDB capture and manage this type of information without heavy customization.

Discovery tools can also register false positives. For example, just because there is a directory on a server labeled “Application X” doesn’t mean that application still runs on that server, and just because you detect traffic on a protocol or port that’s common for a certain application doesn’t mean it’s actually that application.

Tax Audit

“We looked at some of the big vendor CMDB packages but they were overkill and would have required additional resources just to keep them up to date.”

– Matt Estes,
Director of Strategic Initiatives

 
How do you achieve a CMDB that’s accurate and trusted? Here’s how we do it:

  1. Leverage the configuration management data you already have (whether that’s an existing CMDB, discovery tools, an asset management tool, spreadsheets, etc.)
  2. Bring that information together visually in one place (importing or federating the details as necessary)
  3. Leverage modern social collaboration principles to follow and peer review your existing data – so it’s validated and trusted
  4. Fill in the gaps in your existing data with what’s known by your experts (leveraging familiar, easy-to-use social collaboration methods and intuitive visualization of relationships and dependencies)
  5. Use social collaboration to bring experts together to assess change risk, leveraging the information in your CMDB and empowering them to correct any bad information or fill in gaps in real-time as part of their change planning work
  6. Tie in your discovery data on an ongoing basis to validate changes were executed as planned

Working With Multiple CMDBs

ITinvolve federates information from all CMDBs and other data sources, and combines this information with crowd sourcing of your expert team members to truly give you one source of truth about your environment.

We can even work with your existing CMDBs bi-directionally, so their accuracy and completeness is enhanced.

Watch the How to Videos »

 

In 30 days…
You can have an accurate and trusted CMDB that finally delivers the benefits your CMDB was supposed to provide.