Huanhuan Xiong

Migration to PaaS clouds - Migration process and architectural concerns

In the cloud computing technology stack, infrastructure has matured more than platform or software service technologies with respect to languages and techniques used for architecting and managing respective applications. Platform-as-a-Service (PaaS) emerges as a focus for the near future that we will focus on. We look at software architecture and programming concerns in the context of migration to PaaS solutions, i.e. the transition of platform systems from on-premise to cloud solutions. We investigate best-practice approaches in cloud-aware coding in the form of patterns and formulate these as a migration process. While one-to-one mappings of software from on-premise to cloud platforms are possible, statelessness and data externalisation from stateful sessions and applications emerge as solutions if cloud benefits such as elasticity and performance are aimed at.

A Comparison of On-Premise to Cloud Migration Approaches

While cloud computing has certainly gained attention, the potential for increased uptake of the technology is still large. As a consequence, how to move and migrate to the cloud is an unanswered question for many organisations. Gaining an understanding of cloud migration processes from on-premise architectures is our aim here. For this purpose, we look at three provider-driven case studies based on the common three layers of cloud computing: Infrastructure (IaaS), platform (PaaS) and software (SaaS) as a service. These shall be complemented by a fourth, independent systems integration perspective. We extract common migration process activities for the layer-specific processes and discuss commonalities, differences and open issues. The results presented are based on expert interviews and focus groups held with major international cloud solution providers and independent consultants.

CLOUDLIGHTNING: A Framework for a Self-organising and Self-managing Heterogeneous Cloud

As clouds increase in size and as machines of different types are added to the infrastructure in order to maximize performance and power efficiency, heterogeneous clouds are being created. However, exploiting different architectures poses significant challenges. To efficiently access heterogeneous resources and, at the same time, to exploit these resources to reduce application development effort, to make optimisations easier and to simplify service deployment, requires a re-evaluation of our approach to service delivery. We propose a novel cloud management and delivery architecture based on the principles of self-organisation and self-management that shifts the deployment and optimisation effort from the consumer to the software stack running on the cloud infrastructure. Our goal is to address inefficient use of resources and consequently to deliver savings to the cloud provider and consumer in terms of reduced power consumption and improved service delivery, with hyperscale systems particularly in mind. The framework is general but also endeavours to enable cloud services for high performance computing. Infrastructure-as-a-Service provision is the primary use case, however, we posit that genomics, oil and gas exploration, and ray tracing are three downstream use cases that will benefit from the proposed architecture.

Scalable Architectures for Platform-as-a-Service Clouds: Performance and Cost Analysis

Scalability is a significant feature of cloud computing, which addresses to increase or decrease the capacities of allocated virtual resources at application, platform, database and infrastructure level on demand. We investigate scalable architecture solutions for cloud PaaS that allow services to utilize the resources dynamically and effectively without directly affecting users. We have implemented scalable architectures with different session state management solutions, deploying an online shopping cart application in a PaaS solution, and measuring the performance and cost for three server-side session state providers: Caching, SQL database and NoSQL database. A commercial solution with its supporting state management components has been used. Particularly when re-architecting software for the cloud, the trade-off between performance, scalability and cost implications needs to be discussed.

A Generic Framework Supporting Self-Organisation and Self-Management in Hierarchical Systems

A novel, generic, framework for supporting self-organisation and self-management in hierarchical systems is presented. The framework allows for the incorporation of local self-organising and self-managing strategies at each level in the hierarchy. These local strategies determine the behaviour of that level and the effects of these strategies can be communicated to, and used by, the strategies in adjacent levels of the hierarchy. Thus, in general, strategies may be viewed as parameterised functions. Information emanating from both the lower and the upper levels in the hierarchy can be used as parameters. Information from below represents the status of the lower levels, whereas information from above can be used to influence the direction and the rate of system evolution. As the component parts of the system evolve to their goal states, the rate of evolution slows. At that point, by definition, a component is maximally contributing to the global goal state of the system as a whole. A novel concept to measure the distance that a component is from this stasis, its Suitability Index is presented and formally defined. Although the proposed framework can be generalised to any hierarchical system, this paper applies it specifically to large scale, hierarchically structured, computer systems. An implementation of this framework and an empirical study of its effectiveness has been conducted as part of the the CloudLightning Project.

A Database-Specific Pattern for Multi-cloud High Availability and Disaster Recovery

High availability and disaster recovery (HADR) are often discussed in highly critical business systems for business function recovery and continuity concerns. With the development of cloud computing, virtual cloud services are perfectly matched to HADR scenarios, and interoperability is a significant aspect to help users to use HADR service across different cloud platforms and providers. In this paper, we present an architectural pattern describing the integration of high availability and disaster recovery. We focus on database cluster replication between private cloud and public cloud environments. This HADR pattern for database cluster replication implements both synchronous and asynchronous replication concurrently for high availability and disaster recovery purposes. To evaluate the effectiveness of this pattern, we simulate a MySQL-database-cluster HADR scenario under three strategies: hot standby, warm standby and cold standby, and analyze the performance, business continuity features and cost.

Artificial Neural Network Simulation on CUDA

The advent of low cost GPU hardware and user friendly parallel programming APIs, such as NVIDIA CUDA means that affordable, programmable, high-performance computing environments for simulation are now attainable for development of scientific simulations. In this paper the authors present the Mine Hunter program, a parallel simulation of neural networks on NVIDIA CUDA. The simulation consists of 128 mine hunters in a mine field of 8192 mines, running on an Intel Quad Core i5-2500 3.3GHz 2 x Nvidia GeForce GTX 480. The results presented demonstrate that CUDA improves performance by up to 80% compared with the equivalent CPU implementation.

Energy-Efficient Servers and Cloud

As the sizes of cloud infrastructures continue to grow, the complexity of the cloud is becoming more and more difficult to manage. Currently, centralised management schemes dominate and there are already signs that these are no longer fit for purpose. The CloudLightning project takes a novel route, making use of self-organisation techniques to address the problems emerging from the confluence of issues in the emerging cloud: rising complexity and energy costs, problems of management and efficiency of use, the need to efficiently deploy services to a growing community of non-specialist users and the need to facilitate solutions based on heterogeneous components. CloudLightning efficiently addresses three main challenges in the domain of heterogeneous cloud computing: energy efficiency, improved accessibility to cloud and support for heterogeneity. The chapter provides an overview of the CloudLightning system.

Cloud Architectures and Management Approaches

An overview of the traditional three-layer cloud architecture is presented as background for motivating the transition to clouds containing heterogeneous resources. Whereas this transition adds many important features to the cloud, including improved service delivery and reduced energy consumption, it also results in a number of challenges associated with the efficient management of these new and diverse resources. The CloudLightning architecture is proposed as a candidate for addressing this emerging complexity, and a description of its components and their relationships is given.

Self-Organising, Self-Managing Frameworks and Strategies

A novel, general framework that can be used for constructing a self-organising and self-managing system is introduced. This framework is independent of the application domain. It embodies directed evolution, can be parameterised with different strategies, and supports both local and global goals. This framework is then used to apply the principles of self-organisation and self-management to resource management within the CloudLightning architecture.