Randee Adams

Virtualized Infrastructure Impairments

This chapter considers the resource facing service impairments experienced by a cloudbased application for virtualized compute, memory, storage, and networking. These impairments are: VM Failure; Nondelivery of Configured VM Capacity; Delivery of Degraded VM Capacity; ?>Tail ?> Latency; Clock Event Jitter; Clock Drift; and Failed or Slow Allocation and Startup of VM Instance. The chapter begins with a discussion of service latency, virtualization, and cloud computing. Virtualization and cloud add additional service latency risks discussed includes virtualization and cloud causes of latency variation and virtualization overhead and increased variability of infrastructure performance. As the popularity of virtualization and cloud computing grows, infrastructure hardware and software suppliers will focus on improving the performance of their virtualized infrastructure offerings. These improvements, along with the benefits of Moores Law, will improve the quality and consistency of virtualized compute, memory, storage, and networking delivered to applications ?> virtual machine instances.

End-to-End Considerations

End users experience cloud-based applications via their smartphone, tablet, laptop, or other device, and that experience aggregates the service impairments of the cloud-based application, the cloud infrastructure, the wide area and access networks, and the users device itself. This chapter first talks about ?>End-to-End Service Context, ?> where it frames the context and general consideration for service quality actually experienced by end users. Next, it discusses ?>Three-Layer End-to-End Service Model, ?> offering a simple model for analyzing end-to-end service impairments. The section on ?>Distributed Cloud Data Centers, ?> considers the service quality implications of obtaining service from smaller, nearby cloud data centers compared with larger but more distant regional data centers. The chapter also considers complex solutions that rely on resources in multiple cloud data centers. Finally, the chapter talks about ?>Disaster Recovery and Geographic Redundancy, ?> where it considers georedundancy and disaster recovery of cloudbased applications.

Design for Reliability of Cloud Solutions

This chapter contains sections titled: Solution Design for Reliability Solution Scope and Expectations Reliability Requirements Solution Modeling and Analysis Element Reliability Diligence Solution Testing and Validation Track and Analyze Field Performance Other Solution Reliability Diligence Topics

Case Study of Client-Initiated Recovery

This chapter contains sections titled: Overview of DNS Mapping DNS onto Practical Client-Initiated Recovery Model Estimating Input Parameters Predicted Results Discussion of Predicted Results ]]>

Understanding Recovery Timing Parameters

This chapter contains sections titled: Detecting Implicit Failures Understanding and Optimizing RTO