Regrouping metric-space search index for search engine size adaptation

This work contributes to the development of search engines that self-adapt their size in response to fluctuations in workload. Deploying a search engine in an Infrastructure as a Service (IaaS) cloud facilitates allocating or deallocating computational resources to or from the engine. In this paper, we focus on the problem of regrouping the metric-space search index when the number of virtual machines used to run the search engine is modified to reflect changes in workload. We propose an algorithm for incrementally adjusting the index to fit the varying number of virtual machines. We tested its performance using a custom-build prototype search engine deployed in the Amazon EC2 cloud, while calibrating the results to compensate for the performance fluctuations of the platform. Our experiments show that, when compared with computing the index from scratch, the incremental algorithm speeds up the index computation 2–10 times while maintaining a similar search performance.

CloudMPL:a domain specific language for describing management policies for an autonomic cloud infrastructure

To benefit from the advantages that Cloud Computing brings to the IT industry, management policies must be implemented as a part of the operation of the Cloud. Among others, for example, the specification of policies can be used for the management of energy to reduce the cost of running the IT system or also for security policies while handling privacy issues of users. As cloud platforms are large, manual enforcement of policies is not scalable. Hence, autonomic approaches for management policies have recently received a considerable attention. These approaches allow specification of rules that are executed via rule-engines. The process of rules creation starts by the interpretation of the policies drafted by high-rank managers. Then, technical IT staff translate such policies to operational activities to implement them. Such process can start from a textual declarative description and after numerous steps terminates in a set of rules to be executed on a rule engine. To simplify the steps and to bridge the considerable gap between the declarative policies and executable rules, we propose a domain-specific language called CloudMPL. We also design a method of automated transformation of the rules captured in CloudMPL to the popular rule-engine Drools. As the policies are changed over time, code generation will reduce the time required for the implementation of the policies. In addition, using a declarative language for writing the specifications is expected to make the authoring of rules easier. We demonstrate the use of the CloudMPL language into a running example extracted from a management energy consumption case study.

Self-adapting parallel metric-space search engine for variable query loads

This research focuses on automatically adapting a search engine size in response to fluctuations in query workload. Deploying a search engine in an Infrastructure as a Service (IaaS) cloud facilitates allocating or deallocating computer resources to or from the engine. Our solution is to contribute an adaptive search engine that will repeatedly re-evaluate its load and, when appropriate, switch over to a dierent number of active processors. We focus on three aspects and break them out into three sub-problems as follows: Continually determining the Number of Processors (CNP), New Grouping Problem (NGP) and Regrouping Order Problem (ROP). CNP means that (in the light of the changes in the query workload in the search engine) there is a problem of determining the ideal number of processors p active at any given time to use in the search engine and we call this problem CNP. NGP happens when changes in the number of processors are determined and it must also be determined which groups of search data will be distributed across the processors. ROP is how to redistribute this data onto processors while keeping the engine responsive and while also minimising the switchover time and the incurred network load. We propose solutions for these sub-problems. For NGP we propose an algorithm for incrementally adjusting the index to t the varying number of virtual machines. For ROP we present an ecient method for redistributing data among processors while keeping the search engine responsive. Regarding the solution for CNP, we propose an algorithm determining the new size of the search engine by re-evaluating its load. We tested the solution performance using a custom-build prototype search engine deployed in the Amazon EC2 cloud. Our experiments show that when we compare our NGP solution with computing the index from scratch, the incremental algorithm speeds up the index computation 2{10 times while maintaining a similar search performance. The chosen redistribution method is 25% to 50% faster than other methods and reduces the network load around by 30%. For CNP we present a deterministic algorithm that shows a good ability to determine a new size of search engine. When combined, these algorithms give an adapting algorithm that is able to adjust the search engine size with a variable workload.