Anca I. D. Bucur

The performance of processor co-allocation in multicluster systems

In systems consisting of multiple clusters of processors interconnected by relatively slow communication links, co-allocation may be required. We study its performance by means of simulations, depending on the structure and sizes of jobs, and the communication speed ratio. We model a multicluster with C clusters of identical processors. The workload consists of rigid jobs that require fixed numbers of processors, possibly in multiple clusters simultaneously. A job is represented by a tuple of C values, each generated from a same distribution D. In an ordered request the positions of the components in the tuple specify the clusters from which the processors must be allocated. For an unordered request, by the components of the tuple the job only specifies the numbers of processors needed in the separate clusters. A flexible request specifies the total number of processors, obtained as the sum of the values in the tuple. For total requests, there is a single cluster and a request only specifies the total number of processors needed. All intracluster communication links have the same speed, as do all intercluster links. The communication speed ratio is the ratio between the time needed to complete a send operation between processors in different clusters and in the same cluster.

Scheduling Policies for Processor Coallocation in Multicluster Systems

Building multicluster systems out of multiple, geographically distributed clusters interconnected by high-speed wide-area networks can provide access to a larger computational power and to a wider range of resources. Jobs running on multiclusters and, more generally, in grids, may require (processor) coallocation, i.e., the simultaneous allocation of resources (processors) in different clusters or subsystems of a grid. In this paper, we propose four scheduling policies for processor coallocation in multiclusters, and we assess with simulations their performance under a wide variety of parameter settings. In particular, in our simulations we use synthetic workloads and workloads derived from the logs of actual systems and from runtime measurements. We conclude that although coallocation makes scheduling more difficult and the wide-area communication critically impacts the performance, there is a wide range of realistic applications that may benefit from coallocation. However, unrestricted coallocation is not recommended: Limiting the total job size or the number or the sizes of their components improves performance.

A Measurement-Based Simulation Study of Processor Co-allocation in Multicluster Systems

In systems consisting of multiple clusters of processors interconnected by relatively slow network connections such as our Distributed ASCI Supercomputer (DAS), applications may benefit from the availability of processors in multiple clusters. However, the performance of single-application multicluster execution may be degraded due to the slow wide-area links. In addition, scheduling policies for such systems have to deal with more restrictions than schedulers for single clusters in that every component of a job has to fit in separate clusters. In this paper we present a measurement study of the total runtime of two applications, and of the communication time of one of them, both on single clusters and on multicluster systems. In addition, we perform simulations of several multicluster scheduling policies based on our measurement results. Our results show that in many cases, restricted forms of co-allocation in multiclusters have better performance than not allowing co-allocation at all.

The Influence of the Structure and Sizes of Jobs on the Performance of Co-allocation

Over the last decade, much research in the area of scheduling has concentrated on single cluster systems. Less attention has been paid to multicluster systems, although they are gaining more and more importance in practice. We propose a model for scheduling rigid jobs consisting of multiple components in multicluster systems by pure space sharing, based on the Distributed ASCI Supercomputer. Using simulations, we asses the influence of the structure and sizes of the jobs on the systems performance, measured in terms of the average response time and the maximum utilization. We consider three types of requests, total requests, unordered requests and ordered requests, and compare their effect on the systems performance for two scheduling policies, First Come First Served, and Fit Processors First Served, which allows the scheduler to look further in the queue for jobs that fit. These types of job requests are differentiated by the restrictions they impose on the scheduler and by the form of co-allocation used. The results show that the performance improves with decreasing average job size and when fewer restrictions are imposed on the scheduler.

Trace-based simulations of processor co-allocation policies in multiclusters

In systems consisting of multiple clusters of processors which employ space sharing for scheduling jobs, such as our Distributed ASCI (Advanced School for Computing Imaging) Supercomputer (DAS), co-allocation, i.e., the simultaneous allocation of processors to single jobs in multiple clusters, may be required. In this paper we study the performance of several scheduling policies for co-allocating unordered requests in multiclusters with a workload derived from the DAS. We find that beside the policy, limiting the total job size significantly improves the performance, and that for a slowdown of jobs due to global communication bounded by 1.25, co-allocation is a viable choice.

The maximal utilization of processor co-allocation in multicluster systems

In systems consisting of multiple clusters of processors which employ space sharing for scheduling jobs, such as our distributed ASCI supercomputer (DAS), co-allocation, i.e., the simultaneous allocation of processors to single jobs in multiple clusters, may be required. In studies of scheduling in single clusters it has been shown that the achievable (maximal) utilization may be much less than 100%, a problem that may be aggravated in multicluster systems. In this paper we study the maximal utilization when co-allocating jobs in multicluster systems, both with analytic means (we derive exact and approximate formulas when the service-time distribution is exponential), and with simulations with synthetic workloads and with workloads derived from the logs of actual systems.

The Influence of Communication on the Performance of Co-allocation

In systems consistingof multiple clusters of processors interconnected by relatively slow connections such as our Distributed ASCI1 Supercomputer (DAS), jobs may request co-allocation, i.e., the simultaneous allocation of processors in different clusters. The performance of co-allocation may be severely impacted by the slowintercluster connections, and by the types of job requests. We distinguish different job request types ranging from ordered requests that specify the numbers of processors needed in each of the clusters, to flexible requests that only specify a total. We simulate multicluster systems with the FCFS policy-- and with two policies for placinga flexible request, one tries to balance cluster loads and one tries to fill clusters completely--to determine the response times under workloads consistingof a single or of different request types for different communication speeds across the intercluster connections. In addition to a synthetic workload, we also consider a workload derived from measurements of a real application on the DAS. We find that the communication speed difference has a severe impact on response times, that a relatively small amount of capacity is lost due to communication, and that for a mix of request types, the performance is determined not only by the separate behaviours of the different types of requests, but also by the way in which they interact.

Local versus Global Schedulers with Processor Co-allocation in Multicluster Systems

In systems consisting of multiple clusters of processors which employ space sharing for scheduling jobs, such as our Distributed ASCI Supercomputer (DAS), co-allocation, i.e., the simultaneous allocation of processors to single jobs in different clusters, may be required. We study the performance of co-allocation by means of simulations for the mean response time of jobs depending on a set of scheduling decisions such as the number of schedulers and queues in the system, the way jobs with different numbers of components are distributed among these queues and the priorities imposed on the schedulers, and on the composition of the job stream.

Semantic Normalization and Query Abstraction Based on SNOMED-CT and HL7: Supporting Multicentric Clinical Trials

Advances in the use of omic data and other biomarkers are increasing the number of variables in clinical research. Additional data have stratified the population of patients and require that current studies be performed among multiple institutions. Semantic interoperability and standardized data representation are a crucial task in the management of modern clinical trials. In the past few years, different efforts have focused on integrating biomedical information. Due to the complexity of this domain and the specific requirements of clinical research, the majority of data integration tasks are still performed manually. This paper presents a semantic normalization process and a query abstraction mechanism to facilitate data integration and retrieval. A process based on well-established standards from the biomedical domain and the latest semantic web technologies has been developed. Methods proposed in this paper have been tested within the EURECA EU research project, where clinical scenarios require the extraction of semantic knowledge from biomedical vocabularies. The aim of this paper is to provide a novel method to abstract from the data model and query syntax. The proposed approach has been compared with other initiatives in the field by storing the same dataset with each of those solutions. Results show an extended functionality and query capabilities at the cost of slightly worse performance in query execution. Implementations in real settings have shown that following this approach, usable interfaces can be developed to exploit clinical trial data outcomes.

Priorities among multiple queues for processor co-allocation in multicluster systems

In systems consisting of multiple clusters of processors which employ space sharing for scheduling jobs, such as our Distributed ASCI Supercomputer (DAS), co-allocation, i.e., the simultaneous allocation of processors to single jobs in multiple clusters, may be required. In order to handle both single-cluster (local) jobs and multi-cluster (global) jobs, such systems may have only local schedulers (which then need to be aware of the whole system), or only a single global scheduler or both, and each scheduler has its own queue. We assess with simulations the response times of both local and global jobs in multicluster systems for different configurations of queues, for different priority orders in which the associated schedulers are allowed to schedule jobs, and for different job-stream compositions.