Configuration Management: How to really build a CMDB with an auto discovery tool

Service management projects that use a Configuration Management Database (CMDB) as central pointer to enterprise IT asset data should double in number over the next year; an independent research has concluded (Tertio Service Management Solutions). Configuration Management Database uptake is expected to double from 20% to 39% within a year, said the vendor behind the research. The reason for the increased popularity is down to a desire to adopt an ITIL based framework among 44% of respondents.

According to ITIL best practice recommendations, the CMDB should provide accurate information on the configuration of IT assets and their documentation in a way that will support all the other service management processes. For 70% of the organizations polled online during Tertio's research, the perceived benefits that results from implementation of a CMDB is expected to be an improved and coordinated picture of the application infrastructure, and the overall better management of IT assets.

Many companies who deployed Configuration Management have associated this process to the Asset/Inventory Management process/systems for the following reasons:

- Desktop Inventory is often based on inventory existing solutions
- Server Inventory is usually done with System Management solutions such as Tivoli, CA Unicenter, BMC Patrol, etc…

However, there is one activity missing from a Configuration Management view: the building of relationships based on Cis (configuration Items).

To be successful at Business Service Management, companies first should map applications and Services. This is probably the only way to know which IT failure is mucking up Business Services.

To address the issues with the Configuration Management process, companies should find a solution which would avoid creating manually the relationships between the various CIs (Configuration Items) which are identified with the various Inventory tools.

Companies need tools that collect systems configuration data, decode packets and watch kernel I/O to determine application dependencies. These are not service, performance or status monitors rather, they map all the servers, systems, applications, processes, services, users and sometimes even network devices to decipher what applications depend on, and what depends on them. This mapping also makes diagnostic, capacity and Service Management application delivery easier.

These tools should be able to:

- Improve the quality of data in a CMDB
- Be in real-time
- Populate data into the Service Desk
- Avoid to maintain manually the systems
- Keep in real time the relationship between CIs
- Improve all ITIL processes, but specifically Incident Management and Change Management
- Provide Change Impact Analysis features
- Track people changing the infrastructure into production
- Deliver Business Services views
- Map ITIL processes
- Integrate smoothly with both the Service Desk and the System Management solution
- Replicate these relationships into Business Service Management solutions
- Send alerts to the System Management console when for example a change on the Infrastructure is done in production
- Replicate these relationships with Enterprise Architecture platforms such as Mega, Troux technologies, Casewise, Popkin, etc..
- Provide out of the box connectors
- Cope with the long term strategy around Service Management and Enterprise Architecture


There used to be more that 10 to 15 vendors in that space, half of them have been acquired by companies such as Mercury, IBM, Symantec, BMC and others who integrated these tools into their Service Management products.

Companies haven’t yet considered implementing such solutions but this is slowly starting. Auto discovery tools are a must!