The next step is to construct a gap analysis. To do so, the organization must describe current data center practices to identify what capabilities or functions will be needed to meet stated requirements, and to determine if the SCM suites under consideration have such capabilities.
For instance, a requirement may be to standardize server configurations across a data center. But if the current reality is that the organization has allowed a hodgepodge of OS versions, patch states and configurations, the gap analysis will point to a variety of functions needed to meet the requirement of standard configurations, including a discovery capability, automated change tracking and role-based access controls.
The gap analysis may also reveal areas where the SCM product can't meet stated requirements. In such cases, the SCM app must interface with other management systems, such as those needed to map servers and apps to the IT services the company delivers, or to address service/fault monitoring. And while BladeLogic's and Opsware's offerings act as CMDBs (configuration management databases) for servers, IT may want to federate that data store with a generalized CMDB that tracks a wide variety of assets.
A gap analysis also will help you sift through myriad features with which SCM vendors will attempt to woo you. While getting acquainted with the software, our fictional company must rigorously separate the must-haves from the nice-to-haves. Also important as product selection begins is an understanding of the product's complexity. For example, will reports require complex SQL development? Does the distribution sequence for certain software require custom scripting?
Lastly, the description of current data center practices can also serve as baseline metrics against which to measure the SCM system's impact on operational efficiency and provide cost justification for the project. If the average MTTR without SCM is three hours, the IT department should anticipate an improvement in that figure post-deployment. Or, if two FTEs (full-time employees) were dedicated to basic tasks such as server provisioning and software upgrades, IT can measure employee time saved through automation.
