Cloudware Operations and Management - Guide to Cloud Computing for Business and Technology Managers - page 381

time, and/or cost optimization scheduling. It is one of the most
important but also the most difficult parameter to monitor and
regulate as it involves comparison of costing data for exist-
ing services to newly available services as well. Costing data
are complicated and may involve multiple subparameters like
fixed one-time costs and variable costs depending on time and
demand.

b. Regulatory criteria-driven requirements : Regulatory requirements

that need to be fulfilled by the cloud user are further taken over
to the cloud service provider; examples of such data are geo-
graphical constraints like location of data, logging, audit trails,
a nd pr ivac y.

c. Maximum tolerable latency : This is typically a technical metrics,
but in case a business process has some specific response time-
related constraint, then it forms part of the business parameter.

d. Business SLA requirements, including availability guarantees,
availability penalties, and penalty exclusions.

e. Data-related requirements, for example, data security, data loca-
tion, backup, and storage.

f. Role-based access control : These are business-related access control

metrics pertaining to hierarchy of the user organization.

g. On-demand business process deployment and user or applica-
tion constraint management like management of budget and
deadline.

h. Governance criteria like quality of service, policy enforcement,
and management.

i. Environmental constraints, for example, preference for green
providers and incentives for green providers.

2. Technical metrics

a. Virtual machine requirements, say, those related to central pro-
cessing unit (CPU), memory, and operating system

b. Security-related requirements like details related to encryption,
digital key management, and identity and access management

c. Technical SLA requirements, including alerts and notifications
when SLA terms reach a critical point

d. Maximum provisioning time

e. Redundancy and disaster recovery metrics for critical applica-
tions or functionality

f. Environment safety and protection-related metrics like carbon
emission rate and CPU power efficiency