Information Technology Reference
In-Depth Information
can be outsourced and run on external structure need to be identified. As mentioned
in chapter 14.2.2 traditional companies are reluctant to run critical processes or data
on external infrastructure. Besides security and privacy concerns, the application of
Utility and Cloud Computing for many processes might be impeded by legal regu-
lation determining whether data and processes might be deployed on infrastructure
in different countries. In general, suitable processes that can be run on external
resources are:
• Processes and applications that often have unpredictable peaks for computing
and storage resources. Example of such processes and applications are online
sites where demand for resources might be influenced by different seasons or
marketing activities of the company.
• Processes and applications involving non-critical data, but which have high
storage consume and require high online availability.
• Processes and applications that require HPC.
After suitable processes are identified, in a next step a thorough return on invest-
ment calculation is necessary if usage of external resources indeed results in cost
savings and financial gains.
Also in the case of Utility and Cloud Computing, it is advisable to start with
a pilot implementation. The usage of external resources should be supported by
adjustment of IT governance (see section 14.3.2.1) and policies and by change
management (see section 14.3.2.2).
14.3.1.6 Continuous monitoring and Evaluation
Grids and Clouds once institionalised in a company are like a living organism that
can frequently be adjusted according to changing business needs and technical
developments. To enable adjustment of the Grid, a management cycle needs to be
established for constantly measuring both the performance of the Grid or Cloud and
its strategic fit. Thus, as soon as the Grid or Cloud is established also a monitoring
concept needs to be put in place. The components of the monitoring system are:
Grid and Cloud metrics, monitoring and analysis procedures and decision making
processes. The monitoring of the Grid or Cloud should be from two perspectives:
• Technical performance and
• Strategic alignment and fit.
Technical monitoring involves measurement of technical performance indicators
such as availability, performance and response time as well as problem management
and monitoring. Another rather strategic technical aspect of Grid and Cloud moni-
toring is also the long-term capacity planning and adjustment (Murch 2009). The
monitoring of the strategic alignment involves observation of changing business
needs and evaluation of the necessity for respective change and strategic re-align-
ment of the Grid or Cloud.
