Adam P. Chester

Managing Dynamic Enterprise and Urgent Workloads on Clouds Using Layered Queuing and Historical Performance Models

The automatic allocation of enterprise workload to resources can be enhanced by being able to make what–if response time predictions whilst different allocations are being considered. We experimentally investigate an historical and a layered queuing performance model and show how they can provide a good level of support for a dynamic-urgent cloud environment. Using this we define, implement and experimentally investigate the effectiveness of a prediction-based cloud workload and resource management algorithm. Based on these experimental analyses we: (i) comparatively evaluate the layered queuing and historical techniques; (ii) evaluate the effectiveness of the management algorithm in different operating scenarios; and (iii) provide guidance on using prediction-based workload and resource management.

Predictive and Dynamic Resource Allocation for Enterprise Applications

Dynamic resource allocation has the potential to provide significant increases in total revenue in enterprise systems through the reallocation of available resources as the demands on hosted applications change over time. This paper investigates the combination of workload prediction algorithms and switching policies: the former aim to forecast the workload associated with Internet services, the latter switch resources between applications according to certain system criteria. An evaluation of two well known switching policies – the proportional switching policy (PSP) and the bottleneck aware switching policy (BSP) – is conducted in the context of seven workload prediction algorithms. This study uses real-world workload traces consisting of approximately 3.5M requests, and models a multi-tiered, cluster-based, multi-server solution. The results show that a combination of the bottleneck aware switching policy and workload predictions based on an autoregressive, integrated, moving-average model can improve system revenue by as much as 43%.

Dynamic Resource Allocation in Enterprise Systems

It is common that Internet service hosting centres use several logical pools to assign server resources to different applications, and that they try to achieve the highest total revenue by making efficient use of these resources. In this paper, multi-tiered enterprise systems are modelled as multi-class closed queueing networks, with each network station corresponding to each application tier. In such queueing networks, bottlenecks can limit overall system performance, and thus should be avoided. We propose a bottleneck-aware server switching policy, which responds to system bottlenecks and switches servers to alleviate these problems as necessary. The switching engine compares the benefits and penalties of a potential switch, and makes a decision as to whether it is likely to be worthwhile switching. We also propose a simple admission control scheme, in addition to the switching policy, to deal with system overloading and optimise the total revenue of multiple applications in the hosting centre. Performance evaluation has been done via simulation and results are compared with those from a proportional switching policy and also a system that implements no switching policy. The experimental results show that the combination of the bottleneck-aware switching policy and the admission control scheme consistently outperforms the other two policies in terms of revenue contribution.

A System for Dynamic Server Allocation in Application Server Clusters

Application server clusters are often used to service high-throughput web applications. In order to host more than a single application, an organisation will usually procure a separate cluster for each application. Over time the utilisation of the clusters will vary, leading to variation in the response times experienced by users of the applications. Techniques that statically assign servers to each application prevent the system from adapting to changes in the workload, and are thus susceptible to providing unacceptable levels of service. This paper investigates a system for allocating server resources to applications dynamically, thus allowing applications to automatically adapt to variable workloads. Such a scheme requires meticulous system monitoring, a method for switching application servers between \text it {server pools} and a means of calculating when a server switch should be made (balancing switching cost against perceived benefits). Experimentation is performed using such a switching system on a Web application testbed hosting two applications across eight application servers. The testbed is used to compare several theoretically derived switching policies under a variety of workloads. Recommendations are made as to the suitability of different policies under different workload conditions.

Dynamic Resource Allocation for Multi-Tiered, Cluster-Based Web Hosting Environments

e-Business applications are subject to significant variations in workload and this can cause exceptionally long response times for users, the timing out of client requests and/or the dropping of connections. One solution is to host these applications in virtualised server pools, and to dynamically reassign compute servers between pools to meet the demands on the hosted applications. Switching servers between pools is not without cost, and this must therefore be weighed against possible system gain. We combine the reactive behaviour of two switching policies—the Proportional Switching Policy and the Bottleneck Aware Switching Policy—with the proactive properties of seven several workload forecasting models. As each forecasting scheme has its own bias, we also develop four meta-forecasting algorithms to ensure consistent and improved results. We test these schemes with real-world workload traces from several sources and show improvements in servicing capability of up to 40% over real-world Internet traces and 103% over a workload containing extreme events (e.g. flash crowds).

A Framework for Data Center Scale Dynamic Resource Allocation Algorithms

The scale and complexity of online applications and e-business infrastructures has led service providers to rely on the capabilities of large-scale hosting platforms, i.e., data centers. Dynamic resource allocation (DRA) algorithms have been shown to allow server resource allocation to be matched with application workload, which can improve server resource utilisation and drive down costs for service providers. However, research on DRA algorithms has almost exclusively focused on their performance characteristics at small-scale, precluding their useful application in commercial hosting environments, such as those dedicated to supporting cloud computing. In this paper, we show, and subsequently propose a framework to address, the scalability problems of current DRA algorithms. We then build on the proposed framework to develop a highly-scalable algorithm for dynamic resource allocation.

Dynamic Active Window Management: A Method for Improving Revenue Generation in Dynamic Enterprise Systems

In dynamic resource allocation systems, servers are moved between pools when overloading is detected. In this work, we investigate the impact to such systems of combining three adaptive monitoring techniques. First we employ two well known switching policies -- the Proportional Switching Policy (PSP) and the Bottleneck Aware Switching Policy (BSP) -- to move servers between server pools as appropriate. Second we use a meta-forecasting technique to predict the movement in future system workload. Third, we use a Dynamic Active Window Model (DAWM), which defines the period over which workload data is analysed. We have previously shown that request servicing capability can be improved by as much as 40\% when the right combination of dynamic server switching and workload forecasting are used. This extended model shows that a further 51.5\% improvement can be achieved when the switching server policy, meta-forecasting and dynamic active window management are employed together over a real-world workload based on Internet traces.