Configuration Management, Without The Agony
Two confusing terms that routinely have to be explained are Configuration Items (CIs) and Attributes. CIs are like data tables and are structured with the necessary fields, which are the attributes, necessary to adequately describe the CI. CIs are things like hardware, software, documentation and so on. An example of attributes for a server CI include model, processor type, vendor, amount of RAM, and so forth. They describe the CI in greater detail.
Imagine creating a CMDB that has five tables with 10 fields each vs. one with a 1,000 tables with an average of 50 fields each. Now, think about all the data entry, controls and auditing necessary to keep that data accurate. Some firms genuinely believe they need to go for the ultimate solution at the outset. Experience shows that going for the end all be all solution at the start isnt a sound approach.
Instead, care must go into designing a data model that supports the process and objectives of the effort. Literally, chant meaningful and manageable over and over. The data must provide clear value and be manageable, or sustainable, on a day-to-day basis over time. To look at it another way, the value of the configuration data must merit the initial and ongoing costs of maintaining the data.
In closing, the message is straight forward start Configuration Management as simple as possible, learn and evolve the process, underlying data model and supporting software. Configuration Management is an extremely important process that provides other processes with the data needed for them to be successful. For this to happen, Configuration Management itself must be implemented successfully and then evolve over time as the needs of the other processes change as well.