Parallel Servers Research Articles

Analysis of network of biserial queues linked with a common server

This paper is an attempt to study the steady state analysis of a complex network of queues in which a common server is linked in series with each of two parallel biserial servers. The arrival and service pattern follows Poisson law. The generating function technique, law of calculus and statistical tools have been used to find the various queue characteristics. The model finds its application in decision making, in process industry, in banking system, in communication networking, in supermarkets, in healthcare centre and in many administrative setups. An algorithm is proposed to find the mean queue length of customers/jobs, variance of queues and average waiting time of customers/jobs. The performance of algorithm is assessed by a numerical illustration.

Analysis of linkage network of queues with a multi stage flowshop scheduling system

In today’s information age society, where activities are highly interdependent and intertwined, sharing of resources and hence waiting in queues is a common phenomenon that occurs in every facet of our lives. Flowshop scheduling with an objective of minimising the makespan is also an important task in manufacturing system. The present paper is an attempt to establish a linkage network of queues consisting of a biserial server and two parallel servers linked to a common server in series with a multistage flow shop scheduling system. The arrival and service pattern follow Poisson law. The completion time of jobs/customers in queue network (phase I of the service) will be the setup time for first machine in flowshop system having ‘m’ machines in series (phase II of the service). A heuristic approach to find optimal or near optimal sequence of jobs processing minimising the total elapsed time with minimum waiting time has been discussed for the queuing-scheduling linkage model. The performance of the algorithm is evaluated by a numerical illustration.

Product-Form Solutions for a Class of Structured Multidimensional Markov Processes

Motivated by queueing systems with heterogeneous parallel servers, we consider a class of structured multi-dimensional Markov processes whose state space can be partitioned into two parts: a finite set of boundary states and a structured multi-dimensional set of states, exactly one dimension of which is infinite. Using separation of variables, we show that the equilibrium distribution, typically of the queue length, can be represented as a linear combination of product forms. For an important subclass of queueing systems, we characterize explicitly the waiting time distribution in terms of mixtures of exponentials.

Behavior‐based botnet detection in parallel

ABSTRACTBotnet has become one major Internet security issue in recent years. Although signature‐based solutions are accurate, it is not possible to detect bot variants in real‐time. In this paper, we propose behavior‐based botnet detection in parallel (BBDP). BBDP adopts a fuzzy pattern recognition approach to detect bots. It detects a bot based on anomaly behavior in domain name service (DNS) queries and transmission control protocol (TCP) requests. With the design objectives of being efficient and accurate, a bot is detected using the proposed five‐stage process, including: (i) traffic reduction, which shrinks an input trace by deleting unnecessary packets; (ii) feature extraction, which extracts features from a shrunk trace; (iii) data partitioning, which divides features into smaller pieces; (iv) DNS detection phase, which detects bots based on DNS features; and (v) TCP detection phase, which detects bots based on TCP features. The detection phases, which consume approximately 90% of the total detection time, can be dispatched to multiple servers in parallel and make detection in real‐time. The large scale experiments with the Windows Azure cloud service show that BBDP achieves a high true positive rate (95%+) and a low false positive rate ( ∼ 3%). Meanwhile, experiments also show that the performance of BBDP can scale up linearly with the number of servers used to detect bots. Copyright © 2013 John Wiley & Sons, Ltd.

Optimal Routing of Fixed Size Jobs to Two Parallel Servers

We consider a heterogeneous two-server system processing fixed size jobs. This includes the scheduling system, where jobs wait in a common queue, and the dispatching system, where jobs are assigned to server-specific queues upon arrival. The optimal policy with respect to the delay in both systems is a threshold policy characterized by a single parameter. In this special case, the scheduling and dispatching systems achieve the same performance with the optimal threshold. The optimal threshold depends on the arrival rate and the service rates. It can be determined by means of dynamic programming, where the required value functions can be evaluated only at the necessary points by means of efficient Monte Carlo simulations. We also give the optimal threshold at three different limits, which yield a simple closed-form expression for a near-optimal threshold. The optimal policy is illustrated and compared with several heuristic dispatching policies.

Cost-Aware Cooperative Resource Provisioning for Heterogeneous Workloads in Data Centers

Recent cost analysis shows that the server cost still dominates the total cost of high-scale data centers or cloud systems. In this paper, we argue for a new twist on the classical resource provisioning problem: heterogeneous workloads are a fact of life in large-scale data centers, and current resource provisioning solutions do not act upon this heterogeneity. Our contributions are threefold: first, we propose a cooperative resource provisioning solution, and take advantage of differences of heterogeneous workloads so as to decrease their peak resources consumption under competitive conditions; second, for four typical heterogeneous workloads: parallel batch jobs, web servers, search engines, and MapReduce jobs, we build an agile system PhoenixCloud that enables cooperative resource provisioning; and third, we perform a comprehensive evaluation for both real and synthetic workload traces. Our experiments show that our solution could save the server cost aggressively with respect to the noncooperative solutions that are widely used in state-of-the-practice hosting data centers or cloud systems: for example, EC2, which leverages the statistical multiplexing technique, or RightScale, which roughly implements the elastic resource provisioning technique proposed in related state-of-the-art work.

From Local to Global Stability in Stochastic Processing Networks Through Quadratic Lyapunov Functions

We construct a generic, simple, and efficient scheduling policy for stochastic processing networks, and provide a general framework to establish its stability. Our policy is randomized and prioritized: with high probability it prioritizes jobs that have been least routed through the network. We show that the network is globally stable under this policy if there exists an appropriate quadratic local Lyapunov function that provides a negative drift with respect to nominal loads at servers. Applying this generic framework, we obtain stability results for our policy in many important examples of stochastic processing networks: open multiclass queueing networks, parallel server networks, networks of input-queued switches, and a variety of wireless network models with interference constraints. Our main novelty is the construction of an appropriate global Lyapunov function from quadratic local Lyapunov functions, which we believe to be of broader interest.

Dynamic Scheduling of a Two-Server Parallel Server System with Complete Resource Pooling and Reneging in Heavy Traffic: Asymptotic Optimality of a Two-Threshold Policy

We consider a dynamic control problem for a parallel server system commonly known as the N-system. An N-system is a two-server parallel server system with two job classes, one server that can serve both classes, and one server that can only serve one class. We assume that jobs within each class arrive according to a renewal process. The random service time of a job has a general distribution that may depend on both the job's class and the server providing the service. Each job independently reneges, or abandons the queue without receiving service, if service does not begin within an exponentially distributed amount of time. The objective is to minimize the expected infinite horizon discounted cost of holding jobs in the system and having customers abandon, by dynamically scheduling waiting jobs to available servers.It is not possible to solve this control problem exactly, and so, we consider an asymptotic regime in which the system satisfies both a heavy traffic and a resource pooling condition. Then, we solve the limiting Brownian control problem, and interpret its solution as a policy in the original N-system. We label the servers and job classes so that server 1 can only serve class 1 and server 2 can serve both classes. The policy we propose has two thresholds. There is one threshold on the total number of jobs in the system, and one threshold on the number of class 1 jobs in the system. These thresholds are used to determine which job class server 2 should serve. We show that this proposed policy is asymptotically optimal within a specified class of admissible policies in the heavy traffic limit, and has the same limiting cost as the Brownian control problem solution.

Many-server Queues with Customer Abandonment: Numerical Analysis of their Diffusion Model

We use a multidimensional diffusion process to approximate the dynamics of a queue served by many parallel servers. Waiting customers in this queue may abandon the system without service. To analyze the diffusion model, we develop a numerical algorithm for computing its stationary distribution. A crucial part of the algorithm is choosing an appropriate reference density. Using a conjecture on the tail behavior of the limit queue length process, we propose a systematic approach to constructing a reference density. With the proposed reference density, the algorithm is shown to converge quickly in numerical experiments. These experiments demonstrate that the diffusion model is a satisfactory approximation for many-server queues, sometimes for queues with as few as twenty servers.

Analysis of Biserial Servers Linked to a Common Server in Fuzzy Environment

The present paper is an attempt to find various characteristics of a queuing network in which two parallel biserial servers are linked to a common server in series under fuzzy environment. Waiting lines or queues are extensively used to analyze the production and service processes exhibiting random variability in arrival times and service times. It is usually assumed that the time between the two consecutive arrivals and servicing time follows a special probability distribution. However, in real world this type of information is obtained using qualitative data and expressed by words like quick, medium and slow rather than the probabilistic values. The  cut approach and fuzzy arithmetic operations are used to estimate the uncertainty associated with the input parameters. The proposed model is illustrated with a numerical illustration.

WAITING TIME ANALYSIS OF MULTI-CLASS QUEUES WITH IMPATIENT CUSTOMERS

In this paper, we study three delay systems where different classes of impatient customers arrive according to independent Poisson processes. In the first system, a single server receives two classes of customers with general service time requirements, and follows a non-preemptive priority policy in serving them. Both classes of customers abandon the system when their exponentially distributed patience limits expire. The second system comprises parallel and identical servers providing the same type of service for both classes of impatient customers under the non-preemptive priority policy. We assume exponential service times and consider two cases depending on the time-to-abandon distribution being exponentially distributed or deterministic. In either case, we permit different reneging rates or patience limits for each class. Finally, we consider the first-come-first-served policy in single- and multi-server settings. In all models, we obtain the Laplace transform of the virtual waiting time for each class by exploiting the level-crossing method. This enables us to compute the steady-state system performance measures.

REDUCING DOWNLOAD TIME THROUGH MIRROR SERVERS

Downloading large files on internet is very time consuming and usually slow process. As internet is expanding and the size of data on internet is also become larger the user is in need of the download approach which reduces the download time. Parallelization can be the motivation in the direction to download time reduction. The most reliable protocol which is most widely used is File Transfer Protocol. On internet files are replicated on number of servers. These duplicate servers are called Mirror servers which are designed to provide reliability and increase availability. In traditional FTP the client tries to download a file from the single servers. The selection of server is done on the basis of distance. Generally the server which is geographically closest to the client is selected for file download. Parallel FTP proposes simultaneous downloads of disjoint file portion from multiple servers. Parallel FTP server selects file server for the requesting client on the basis of variety of QoS parameter such as Available Bandwidth and Server Utilization. We perform a simulation based comparative study between Traditional FTP and Parallel FTP. In our approach we also tried to add some more QoS parameter which are Hop Counts and Delay. The file download time is inversely proportional to number of hops and directly proportional to idleness of servers. The simulation is done on Network Simulator -2.

FCC – An automated rule-based processing tool for life science data

BackgroundData processing in the bioinformatics field often involves the handling of diverse software programs in one workflow. The field is lacking a set of standards for file formats so that files have to be processed in different ways in order to make them compatible to different analysis programs. The problem is that mass spectrometry vendors at most provide only closed-source Windows libraries to programmatically access their proprietary binary formats. This prohibits the creation of an efficient and unified tool that fits all processing needs of the users. Therefore, researchers are spending a significant amount of time using GUI-based conversion and processing programs. Besides the time needed for manual usage, such programs also can show long running times for processing, because most of them make use of only a single CPU. In particular, algorithms to enhance data quality, e.g. peak picking or deconvolution of spectra, add waiting time for the users.ResultsTo automate these processing tasks and let them run continuously without user interaction, we developed the FGCZ Converter Control (FCC) at the Functional Genomics Center Zurich (FGCZ) core facility. The FCC is a rule-based system for automated file processing that reduces the operation of diverse programs to a single configuration task. Using filtering rules for raw data files, the parameters for all tasks can be custom-tailored to the needs of every single researcher and processing can run automatically and efficiently on any number of servers in parallel using all available CPU resources.ConclusionsFCC has been used intensively at FGCZ for processing more than hundred thousand mass spectrometry raw files so far. Since we know that many other research facilities have similar problems, we would like to report on our tool and the accompanying ideas for an efficient set-up for potential reuse.

Many-server queues with customer abandonment: Numerical analysis of their diffusion model

Open AccessOpen Access licenseAboutSectionsView PDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareShare onFacebookTwitterLinked InEmail Go to SectionOpen AccessOpen Access license HomeStochastic SystemsVol. 3, No. 1 Many-server Queues with Customer Abandonment: Numerical Analysis of their Diffusion ModelJ. G. Dai, Shuangchi HeJ. G. Dai, Shuangchi HePublished Online:20 May 2013https://doi.org/10.1287/11-SSY029AbstractWe use a multidimensional diffusion process to approximate the dynamics of a queue served by many parallel servers. Waiting customers in this queue may abandon the system without service. To analyze the diffusion model, we develop a numerical algorithm for computing its stationary distribution. A crucial part of the algorithm is choosing an appropriate reference density. Using a conjecture on the tail behavior of the limit queue length process, we propose a systematic approach to constructing a reference density. With the proposed reference density, the algorithm is shown to converge quickly in numerical experiments. These experiments demonstrate that the diffusion model is a satisfactory approximation for many-server queues, sometimes for queues with as few as twenty servers. Previous Back to Top Next FiguresReferencesRelatedInformationCited BySRPT Scheduling Discipline in Many-Server Queues with Impatient CustomersJing Dong, Rouba Ibrahim21 October 2021 | Management Science, Vol. 67, No. 12Diffusion Approximation for Efficiency-Driven Queues When Customers Are PatientShuangchi He22 May 2020 | Operations Research, Vol. 68, No. 4Dynamic Scheduling in a Many-Server, Multiclass System: The Role of Customer Impatience in Large SystemsJeunghyun Kim, Ramandeep S. Randhawa, Amy R. Ward11 April 2018 | Manufacturing & Service Operations Management, Vol. 20, No. 2Stein’s Method for Steady-state Diffusion Approximations: An Introduction through the Erlang-A and Erlang-C ModelsAnton Braverman, J. G. Dai, Jiekun Feng17 February 2017 | Stochastic Systems, Vol. 6, No. 2A Unified Approach to Diffusion Analysis of Queues with General Patience-Time DistributionsJunfei Huang, Hanqin Zhang, Jiheng Zhang3 March 2016 | Mathematics of Operations Research, Vol. 41, No. 3 Volume 3, Issue 1June 2013Pages 1-321 Article Information Metrics Downloaded 129 times in the past 12 months Information Received:April 01, 2011Published Online:May 20, 2013 Copyright © 2013, The author(s)Cite asJ. G. Dai, Shuangchi He (2013) Many-server Queues with Customer Abandonment: Numerical Analysis of their Diffusion Model. Stochastic Systems 3(1):96-146. https://doi.org/10.1287/11-SSY029 KeywordsDiffusion processstationary distributionphase-type distributionmany-server queueheavy trafficcustomer abandonmentquality- and efficiency-driven regimePDF download

Subspace Snooping: Exploiting Temporal Sharing Stability for Snoop Reduction

Although snoop-based coherence protocols provide fast cache-to-cache transfers with a simple and robust coherence mechanism, scaling the protocols has been difficult due to the overheads of broadcast snooping. In this paper, we propose a coherence filtering technique called subspace snooping, which stores the potential sharers of each memory page in the page table entry. By using the sharer information in the page table entry, coherence transactions for a page generate snoop requests only to the subset of nodes in the system. However, the coherence subspace of a page may evolve, as the phases of applications may change or the operating system may migrate threads to different nodes. To adjust subspaces dynamically, subspace snooping supports two different shrinking mechanisms, which remove obsolete nodes from subspaces. Among the two shrinking mechanisms, subspace snooping with safe shrinking can be integrated to any type of coherence protocols and network topologies, as it guarantees that a subspace always contains the precise sharers of a page. Speculative shrinking breaks the subspace superset property, but achieves better snoop reductions than safe shrinking. We evaluate subspace snooping with Token Coherence on unordered mesh networks. Subspace snooping reduces 58 percent of snoops on average for a set of parallel scientific and server workloads, and 87 percent for our multiprogrammed workloads.

Load balancing via random local search in closed and open systems

In this paper, we analyze the performance of random load resampling and migration strategies in parallel server systems. Clients initially attach themselves to an arbitrary server, but may switch servers independently at random instants of time in an attempt to improve their service rate. This approach to load balancing contrasts with traditional approaches where clients make smart server selections upon arrival (e.g., Join-the-Shortest-Queue policy and variants thereof). Load resampling is particularly relevant in scenarios where clients cannot predict the load of a server before being actually attached to it. An important example is in wireless spectrum sharing where clients try to share a set of frequency bands in a distributed manner. We first analyze the natural Random Local Search (RLS) strategy. Under this strategy, after sampling a new server randomly, clients only switch to it if their service rate is improved. In closed systems, where the client population is fixed, we derive tight estimates of the time it takes under RLS strategy to balance the load across servers. We then study open systems where clients arrive according to a random process and leave the system upon service completion. In this scenario, we analyze how client migrations within the system interact with the system dynamics induced by client arrivals and departures. We compare the load-aware RLS strategy to a load-oblivious strategy in which clients just randomly switch server without accounting for the server loads. Surprisingly, we show that both load-oblivious and load-aware strategies stabilize the system whenever this is at all possible. We use large-system asymptotics to characterize system performance, and augment this with simulations, which suggest that the average client sojourn time under the load-oblivious strategy is not considerably reduced when clients apply smarter load-aware strategies.

A finite source multi-server inventory system with service facility

In this article, we study a continuous review retrial inventory system with a finite source of customers and identical multiple servers in parallel. The customers arrive according a quasi-random process. The customers demand unit item and the demanded items are delivered after performing some service the duration of which is distributed as exponential. The ordering policy is according to (s, S) policy. The lead times for the orders are assumed to have independent and identical exponential distributions. The arriving customer who finds all servers are busy or all items are in service, joins an orbit. These orbiting customer competes for service by sending out signals at random times until she finds a free server and at least one item is not in the service. The inter-retrial times are exponentially distributed with parameter depending on the number of customers in the orbit. The joint probability distribution of the number of customer in the orbit, the number of busy servers and the inventory level is obtained in the steady state case. The Laplace–Stieltjes transform of the waiting time distribution and the moments of the waiting time distribution are calculated. Various measures of stationary system performance are computed and the total expected cost per unit time is calculated. The results are illustrated numerically.

Stability analysis of N-model systems under a static priority rule

We consider the stability of N-model systems that consist of two customer classes and two server pools. Servers in one of the pools can serve both classes, but those in the other pool can serve only one of the classes. The standard fluid models in general are not sufficient to establish the stability region of these systems under static priority policies. Therefore, we use a novel and a general approach to augment the fluid model equations based on induced Markov chains. Using this new approach, we establish the stability region of these systems under a static priority rule with thresholds when the service and interarrival times have phase-type distributions. We show that, in certain cases, the stability region depends on the distributions of the service and interarrival times (beyond their mean), on the number of servers in the system, and on the threshold value. We also show that it is possible to expand the stability region in these systems by increasing the variability of the service times (without changing their mean) while keeping the other parameters fixed. The extension of our results to parallel server systems and general service time distributions is also discussed.

A Read-Copy Update based parallel server for distributed crowd simulations

The Read-Copy Update (RCU) synchronization method was designed to cope with multiprocessor scalability some years ago, and it was included in the Linux kernel October of 2002. Recently, libraries p...

CEA: A Cyclic Expansion Algorithm for data migration in parallel video servers

Parallel video servers can achieve highly storage-saving and granularly load-balancing, but they suffer from a system expansion problem. As the number of users continuously increases, the system inevitably needs to expand the number of video servers. However, the expansion of a parallel video server system is not as simple as that of a replicated video server system. Hence, this work develops an efficient expansion algorithm, called the Cyclic Expansion Algorithm (CEA), for parallel video servers. The proposed CEA algorithm has several good features. First, the data layout of each video content exhibits periodicity. Consequently, the meta-data size of each video and the complexity of the CEA algorithm are reduced. Second, the number of required data movements during a system expansion is optimized. Third, the total number of required XOR recomputations for updating parity blocks during an expansion is also minimized. Additionally, the new CEA can be applied to a variety of distributed storage systems, such as the cloud-based storage systems using striping and parity check techniques.