In general, performance measures characteristics of the higher service layers depend on those of the underlying technology components. Since consumers typically have no visibility into technology specifics, they can be clueless concerning performance expectations.
The typical user operational metrics used to evaluate cloud service providers are the following:
- Service response time (delay): The latency time between service request and service completion
- Service throughput: The number of jobs processed by the service provider within a set time unit
- Service availability: The probability that the service provider accepts a customer service request at any time
- System utilization: The percentage of system resources being used for service provisioning
- System resilience: The stability of system performance over time, especially under bursty loads
- System scalability...
