IT Service Measurement Automation

[JonahG]

I'm curious if some companies that are adopting ITIL have looked at ways of automating their service measurement calcualations. We are exploring the possibility of rolling up the data from our management systems to generate our availability numbers. Are there any papers that speak to established best practices towards this?

[wvasquez]

Best practice is not to try to roll-up components into an overall calculation.

Instead using tools such as BMC's Patrol End-to-End Response Timer for example, the service availability and performance targets can be effectively measured against. This provides a pragmatic and practical approach to reporting on service availability and performance and is considered the best practice approach.

[Darreck]

JonahG wrote:

We are exploring the possibility of rolling up the data from our management systems to generate our availability numbers.

Wvasquez, everything you said is very true and I value your wisdom.

Jonah’s comment above opens up the reoccurring scenario that many businesses fall into when trying to capture valid Availability data to support planning for Capacity.

We have all seen it, “ I have all of these monitoring tools that I am paying big money so create a process to support my tool.” Availability metrics should draw upon its supporting processes, to include financial, configuration, change, and capacity to name a few. It should not solely reside on the raw data being captured from monitors.

With out these processes in place the effectiveness of your Availability process will be sufficient at best.

SLM is going to be the focal point of your metric requirements and reporting. I would look into Service Level Management for guidance.

So the term Best Practices really covers the lessons learned from prior implementations and saving us the time of not repeating them.

Sorry for my sermon but I have relived this on many clients sites.