CMDB: A Long Road to ITIL's Return-On-Investment?

CMDB: A Long Road to ITIL’s Return-On-Investment?
By John Worthington.


A key concept of ITIL’s best practice for IT Service Management is obviously to focus on Services, from the business’ perspective (read business process). So this means that Citrix, Web front-end infrastructures, and even SAP are not technically services, but segments of end-to-end services that make sense to IT (not IT’s customer). Supply chain, logistics, sales accounting, might be more likely to be considered services from the business perspective.

So, ITIL defines core Service Support and Service Delivery processes and suggests roles and responsibilities that can cross organizational silos and begin to treat IT from a services orientation. People, Process and Products (tools) become part of an overall Continuous Service Improvement Program (CSIP).

It does not take long for customers to learn that a key area and challenge will be the establishment of a Configuration Management Data Base (CMDB).

As most ITIL-ers know, the key difference between a CMDB and an asset management system is that the CMDB also contains relationships between Configuration Items (hardware, software, et al).

Establishing the relationships between components for an end-to-end business service is a big challenge. In fact, some tool vendors are selling ‘automated’ dependency mapping as a standalone product.

In fact, the costs associated with CMDB tools can be very significant both in terms of hardware/software and implementation. Placing this cost burden too early in an ITIL deployment can sour management on the overall benefits of the ITIL implementation, with the potential to mortally wound the program before it really gets anything going. This is especially true for tool-driven implementations.

Some customers implementing ITIL best practice have quickly become engrossed in the mud associated with detailed data base design efforts, in an attempt to make sure the CMDB is designed properly and anticipates the entire organizations’ needs.

While CMDB design is important, this can drag out the implementation of ITIL significantly.

In addition, it tends to focus the customers attention towards Change and Configuration Management tools and on service segments, rather than end-to-end business processes. This is the attempt to “eat the elephant in small bites” philosophy. This can delay true business alignment which is a fundamental objective of IT Service Management.

To make matters worse, in many organizations most of the costs of IT are spent in isolation and diagnosis on problems in n-tier infrastructures. Focus on CMDB design and ‘rapid’ implementation of Change and Configuration management, across only some segments of the end-to-end service infrastructure (read silos) - may not help accelerate ROI and business alignment fast enough to meet the needs of the business. In addition, even the most robust Change/Config tool may not isolate a problem. It also has the potential to drive the decision for a CMDB tool sooner than the organization is really ready.

Finally, so much focus on Service Support puts opportunities to leverage Service Delivery processes on a back burner. This almost exclusive focus on Service Support actually compounds the delay in ITIL’s ROI since all ITSM processes are inter-related and achievement of the overall Continuous Service Improvement Program goals depend on ALL the process areas being implemented.

Is there another (faster) road?

While there are no shortcuts to proper design and construction of a CMDB, there may be other opportunities to accelerate the ROI associated with implementing best practice.

I have worked with a partner who has been providing customers with an “EasyEVAL Program” for over a year, as a somewhat traditional approach to evaluating software. Typically, a customer identifies a target segment (Citrix, Web, etc.) of a service and excersizes the software in this environment. The usual software sales pitch…

However, customers could be missing an opportunity to put real VALUE in these eVALUations. I have seen clients quickly establish a services-oriented CDB (not really a CMDB), that contains many of the exact relationships that the CMDB will require. We have done this in some pretty complex environments in less than a week. The file containing these dependencies could be exported to the team responsible for populating the CMDB however our focus has been on establishing cross-silo baselines of performance across an end-to-end business service.

For most customers, the compelling aspect of the technology is a patented data flow and dependency based correlation technology that essentially automates the triage process that occurs as customers attempt to isolate n-tier infrastructure problems, directly addressing a significant source of costs for many IT shops. However, it also greatly has accelerated the IT staff’s orientation from silos to services, via a unique and very simple user interface.

The company’s willingness to conduct proof of concepts across n-tier business services is a real opportunity for customers to quickly obtain a clear definition of the service dependencies and a performance base line across every layer of every tier of a business service.

So, go ahead and take the proper steps to carefully plan your CMDB and choose your vendor carefully. However, in the meantime don’t pass an opportunity to leverage the potential to establish a service oriented CDB, automate the triage process, accelerate the paradigm shift to a service orientation and help discover CMDB application dependencies; especially when this can be done as a POC or small consulting engagement!

Even if the tool is not selected, the result will be a more effective CMDB selection and acceleration of ITIL’s ROI at the same time. I’d be interested in other comments on how performance management tools and/or the CDB is complimenting ITIL implementations.

For more information - click MyServiceMonitor

John M. Worthington
IT Service Management Consulting
MyServiceMonitor, LLC.

John holds both a Manager’s Cert and a Practitioner Cert in Release & Control.