September 25, 2012
If you're at this level, you probably...
- Compile workload information across tiers to analyze and report on services;
- Have established a dialogue with the business, trying to anticipate changes; and
- Use analytical modeling to predict the outcome of those scenarios.
However, you also probably...
- Are not accustomed to reporting IT results in business productivity or financial terms.
The Service level is a continuation of the Proactive level, but with a stronger focus on
workloads representing services. The view is extended to cover the full spectrum of
workloads that make up a particular service, from the end user to the backend systems.
By automatically gathering the information needed from each of the components that
comprise the service, one can detect, prioritize and execute those actions that will
improve end user experience.
To achieve this, simple trending is not enough to predict future needs. When looking
at the complex environment needed to deliver a service, it is necessary to have more
advanced tools that predict the effects of future business changes.
Analytical modeling offers a simple yet powerful way to address different scenarios.
Based on the empirical data collected during the monitoring of systems, models of the
systems are created. These models can then be used to accurately predict the impact
of growth, changes to the infrastructure, migration of workloads, etc.
Reaching this level takes more than a focus on technology. You also
need to establish processes for exchanging information. To identify
and verify prediction scenarios, there must be a two-way
communication between the business and IT. If the business wants IT to provide a
certain level of service without excessive over-provisioning, it needs to supply IT with
growth projections and business plans. On the other hand, to demonstrate the value
of the information received, IT needs to share cost-benefit analyses of this planning
with the business. Building this mutual trust is integral to success.
At this level, since IT knows the characteristics and importance of different services,
they can start to optimize the usage of them by implementing chargeback mechanisms
where different parties pay for the actual use of resources. This will yield not only a
fair distribution of cost, but also a way for IT to influence the usage patterns. Offering
a discount during low activity hours might smooth out the peaks and troughs often
seen in a data center.
See where your organization falls in the TeamQuest Capacity Management Maturity Model