Modeling Parallel and Distributed Systems with Finite Workloads

Size: px
Start display at page:

Download "Modeling Parallel and Distributed Systems with Finite Workloads"

Transcription

1 Modeling Parallel and Distributed Systems with Finite Workloads Ahmed M. Mohamed, Lester Lipsky and Reda Ammar {ahmed, lester, Dept. of Computer Science and Engineering University of Connecticut, Storrs, CT 6269, USA Abstract In studying or designing parallel and distributed systems one should have available a robust analytical model that includes the major parameters that determine the system performance. Jackson networks have been very successful in modeling parallel and distributed systems. However, they have their limitations. In particular, the product-form solution of Jackson networks assumes steady state and exponential service centers or certain specialized queueing disciplines. In this paper, we use a transient model studying distributed systems with finite workload (no new arrivals). Using some nonexponential distributions we show to what extent the exponential distribution can be used to approximate other distributions. When the number of tasks to be executed is large enough, the model approaches the product-form solution in those cases where the Jackson networks can be applied. We also study some cases where Jackson networks can t be applied (the nonexponential servers have queueing). The model can be used for reliability analysis of systems that allow failures without repair (fail-stop). Key words: Analytical Modeling, Performance Prediction, Queueing Models, Jackson Networks Reliability Analysis and Transient Analysis.. Introduction It is often assumed that tasks in parallel systems are independent and run independently using separate hardware (e.g. Fork/Join type applications). In this case the problem is reduced to order statistics [22,23,24,25]. However, in many applications tasks must interact through shared resources (e.g. shared data, communication channels). Hence, order statistics analysis is not adequate and one must apply more general models. We used the product-form solution for Jackson networks in [,2] to model clusters of workstations. This model is satisfactory if the number of tasks is much greater than the degree of multi-tasking and the shared resources do not have high variance for task services. However, if this is not the case, the transient model discussed here should be employed. The steady state solution of Jackson networks assumes exponential service times. It has been shown by Leland and Ott [5] that the distribution of the CPU times at BELLCORE are power tailed (PT). Also, Lipsky and Crovela[6,9], Hatem[3] and others, found that file sizes stored on disks are PT. If these observations are true generally, then performance modeling based on exponential distributions are not adequate. In a previous work, we developed a transient model for Jackson networks for exponential distributions [3]. In [5] we presented a transient model in which the dedicated servers (i.e., local CPU s and Disks) have nonexponential service times. In this case, the steady state value of the model is the same as the steady state value of the product form solution of Jackson networks. In this paper, we present the case when shared servers (i.e., shared disks and the communication network) have nonexponential service times (Jackson networks can t be applied to such systems). The model we present includes the several performance parameters that affect the performance of parallel and distributed systems. More parameters always can be added to the basic model (e.g., scheduling overhead, multitasking,.etc) if needed. Also, the model can be used in reliability analysis for systems that allow failures without repair (fail-stop). In our analysis, we use the Linear Algebraic Queueing Theory (LAQT) approach. All the necessary background needed can be found in [7]. The rest of the paper is organized as follows: In Section 2, we give a brief background. In Section 3, we introduce our transient model. In Section 4, we show how to use the model to analyze the performance of different configurations of parallel and distributed systems. In Section 5, we show how to use the model in reliability analysis. In Section 6, we show some of our results.

2 2. Background Since the early 97s, networks of queues have been studied and applied to numerous areas in computer science and engineering with a high degree of success. General exponential queueing network models were first solved by Jackson [4] and by Gordon and Newell [2] who showed that certain classes of steady state queueing networks with any number of service centers could be solved using a product-form solution. A substantial contribution was made by Buzen [7,8] who showed that the ominouslooking formulas were computationally manageable. Basket et al [2] summarized under what conditions the product form solution could be used (e.g., processor sharing, multiple classes, etc). Thereafter, the analysis of queueing networks began to be considered as a research field of its own. We used the product form solution of Jackson networks to analyze the performance of clusters of workstations []. The model then was used to develop efficient data allocation algorithms [2]. Then we [3] developed a transient model for Jackson networks and showed to what extent the steady state model can be used. However, there is a computational problem (state space explosion) in attempting to use this model for large systems. An approximation to the transient model was presented in [4]. 3. The Transient model First we introduce some definitions that are important for our analysis. A complete description can be found in [6]. - Objects used if only one customer in the system S is a system consisting of a set of service centers. Ξ is the set of all internal states of S. p is the entrance row vector where p i is the probability that upon entering S, a customer will go to server i. q is the exit vector where q i is the probability of leaving the system when service completed at server i. M is the completion rate matrix whose diagonal elements are the completion rates of the individual servers. The rest of the elements are zeros. P is the transition matrix where P ij is the probability that a customer will go from server i to server j when service is completed at i. B is the service rate matrix, B = M (I P). τ is a column vector where τ i is the mean time until a customer leaves S, given that he started at server i. V is the service time matrix where V ij is the mean time a customer spends at j from the time it first visits i until it leaves the system. V = B - ε is a column vector all of whose components are ones. - Objects used if k > tasks in the system Ξ k is the set of all internal states of S when there are k active customers there. There are D(k) such states. M k is the completion rate matrix where [M k ] ii the service rate of leaving state i. The rest of the elements are zeros. P k is the transition matrix where [P k ] ij i,j Ξ k, is the probability that the system will go from state i to state j when service is completed while the system is in state i. Q k is the exit matrix where [Q k ] ij is the probability of a customer leaving S when the system was in state i Ξ k, leaves the system in state j Ξ k-. R k entrance matrix where [R k ] ij is the probability that a customer upon entering S finding it in state i Ξ k- will go to server that puts the system in state j Ξ k. τ k is a column vector of dimension D(k) where [τ k ] i is the mean time until a customer leaves S, given that the system started in state i Ξ k. Suppose we have a computer system made up of K processors or workstations and we wish to compute a job made of N tasks where N > K. The first K tasks are assigned to the system and the rest are queued up waiting for service. Once a task finishes it is immediately replaced by another task from the execution queue. Assume that the system initially opens up and K tasks flow in. The first task enters putting the system in state p Ξ.The second task enters and takes the system from that state to state pr 2 Ξ 2 and so on. The state of the system after the K th task enters is: p K = pr 2 R 3 R K It can be shown that the mean time until the first task finishes and leaves the system is [4], τ K = M k - ε k + P k τ k τ K = (I k P k) - M k - ε k = V k ε k The mean time is given by, t k = p k V k ε k How long does it take for the next tasks to finish?. There are two possibilities, either N = K or N > K. 3. Case. (N = K) First, define matrix Y k where [Y k ] ij is the probability that S will be in state j Ξ k- immediately after a departure, given that the system was in state i Ξ k and no other customers have entered. Y k can be obtained from the following argument. When an event occurs in S, either someone leaves, [Q k ], or the internal state of the system changes, [P k ], and eventually somebody leaves, [Y k ]. Y k = Q k + P k Y k Y k = (I k P k ) - Q k = (I k P k ) - M k - M k Q k =V k M k Q k k K

3 We then consider how long it takes for the second task to finish after the first one left. p K Y K (V K- ε K- ) = p K Y K (τ K- ) where, [p k ] i is the probability that the system was in state i when epoch began (epoch is the time between two successive departure). This means, after the first task leaves, the system is in state p k Y k with k, tasks. The second task takes (τ k- ) to leave next. The time between the second and third departures is, p K Y K (Y K- ) (V K-2 ε K-2 ) = p K Y K (Y K- ) (τ K-2 ) In general the mean time to finish all tasks is given by, E(T) = p K[τ K + Y K τ K Y K Y K - Y τ ] 3.2 Case 2. (N > K) The first K tasks are assigned to the system and the rest are queued up waiting for service. When a task leaves the system, another one immediately takes its place, putting the system in state Y K R K = V K M K Q K R K where, Y K R K ε K = ε K The mean time until the second task finishes is: p K Y K R K (V K ε K ) = p K Y K R K (τ K) Now, another tasks enters the system, putting the system in state, Y K R K Y K R K = (V K M K Q K R K ) 2 The mean time until the third task finishes is given by, p K (Y K R K ) 2 (V K ε K) = p K (Y K R K ) 2 (τ K) Eventually we will reach case one again, where there are only K tasks remaining, but with initial state p K (Y K R K ) N-K. In general, the mean time to finish executing all tasks is, N K E(T) = p K [ (Y K R K ) i ] (τ K) i= + p K (Y K R K ) N-K Y K [τ K + Y K τ K- + Y K Y K- τ K Y K Y K- Y τ ] Each term of the above equation helps describe the transient behavior. When i is small, the term (Y K R K ) i is different for different values of i which gives different departure times for different epochs. Once i becomes large the term approaches ε K p ss (p SS Y K R K = p SS )gives the steady state solution for Jackson networks where applicable. Once the number of tasks remaining becomes less than the number of processors, we have different values of k (k < K) for different system sizes, which leads to the other transient region (draining region). 4. Modeling approach The success of a performance model is dependent on how accurately the model matches the system and more importantly what insights does it provide for performance analysis. The major parameters we are modeling include communication contention, geometry configurations, time needed to access different resources and data distribution. Such a model is useful to get a basic understanding of how the system performs. More details can always be added to the basic model (e.g., scheduler overheads, multitasking, task dependencies, etc) but with increase in the state space. 4. Application model The target parallel application can be considered to be a set of independent, identically distributed (iid) tasks, {t, t 2... t N }, The tasks are queued up (if N > K), and the first K tasks are assigned to the system. When a task is finished, it is immediately replaced by another task in the queue. The active tasks can communicate with each other by exchanging data through each other's disks. The tasks run in parallel, but they must queue up for service when they wish to access the same device. Each task consists of a finite number of I/O and CPU instructions. Therefore, the execution of the task consists of phases of computation, then I/O then computation, etc, until finished. We assume that during an I/O phase the task cannot start a new computational phase (no CPU-I/O overlap). Assume that T is the random variable that represents the running time of a task if it is alone in the system. Then, the mean execution time E(T) for a task can be divided into three components: T, T 2 and T 3, where, E(T ) is the expected time needed to execute non-i/o instructions locally (local CPU time). E(T 2 ) is the expected time needed to execute I/O instructions locally (local disk time) E(T 3 ) is the expected time needed to execute I/O instructions else where(remote disk time) As in [,2], we use the following parameters to represent the above components. X = E(T ) + E(T 2 ), C * X = E(T ), ( C) * X = E(T 2 ), Y = E(T 3 ). C is the fraction of local time spent at the local CPU. So based on the above parameters we can write T as: E(T) = C * X + ( C ) * X + Y. All of this is for the case when there is only one task in the system (therefore no contention). The performance model uses these parameters to calculate the effect of contention when more than one task is running. 4.2 System model It is assumed that the computer system consists of a network of workstations under the control of a single scheduling mechanism with a centralized data storage disk.

4 All workstations contact this central disk when they request global data. 4.3 Task activity Each task spends some time in its local CPU doing computation then, assuming it is not finished, with probability (-q) goes to some disk for I/O. Tasks may need access a remote disk(s) through a communication channel with probability (-q)p 2 and their local disk with probability (-q)p After finishing its I/O the task returns to its local workstation. The task finishes its execution and leave the system with probability q. Thus, the number of computational cycles is geometrically distributed with mean of (/q) cycles. 4.4 Modeling central storage systems The system consists of K workstations and a central server. A workstation can be modeled as one server or more based on the assumptions made. Here, we assume that each workstation consists of two servers (CPU and a disk). The K workstations are connected by a shared network channel (one server). Therefore the number of servers needed to model K workstations is 2*K + 2. The Kronecker-product formulation of the system would require the following number of states: D(K) = (2*K + 2) K Since we assumed that the tasks are iid, we can use a reduced product representation. Even after using the reduced product space the number of states can still be huge. We still need to reduce the number of states. Since the tasks never compete for CPU s or local I/O, an equivalent model to the above description is considered. We can consider the first server in the system to be a load dependent server representing all of the CPUs. Then another load dependent server can represent all of the local disks. Finally two load independent servers represent the communication channel and the central disk. Therefore, the number of servers is four, which reduces the state space to: D RP ( K ) K + = K 3 Assume we are modeling a system with five workstations and all servers have exponential service times. The following are the basic matrices (when only one task is running in the cluster). p = [,,, ], as we assume that a task starts its execution from the CPU. q = [q,,, ], as we assume that a task always leaves from the CPU. CPU Disk Comm. R.Disk P p( q) p ( q ) 2 = M µ cpu = µ d µ com where, p + p 2 = From V = M (I P) - it follows that: pv = [t cpu /q, t d * p (-q)/q,t com *p 2 (-q)/q, t rd *p 2 (-q)/q] where, t cpu = /µ cpu, t d = /µ d, t com = /µ com, t rd = /µ rd µ rd pv is the time components vector that represents the total time spent by a task in each stage of the system (for the case k = ). For example the task will spend (t cpu /q) units of time in its CPU. The M matrix is known from the architecture of the cluster but the reformation of the P matrix to the parameters given in the application model described earlier is still to be specified. The application model specifies the total time spent by a task in the cluster (with no contention) as [CX, (- C)X, BY, Y]. By using these time components we obtain the parameters in the P matrix as follows, q = t cpu / CX p = q * ( C) X / t d * ( q) p 2 = q * Y / t rd * ( q) If we wish to use non-exponential distribution as the service distribution for the CPU, we only need to change the matrices as shown in the next subsection Erlangian distributions The Erlangian-m distribution describes the time it takes for a customer to be served by m identical exponential servers in series. Therefore, the Erlangian- is the exponential distribution and its pdf is f(t) = µ exp(-µ t) where µ is the service rate, t cpui = /µ cpui The pdf of the Erlangian-2 distribution is f(t) = µ (µ t) exp(-µ t) If we wish to use the Erlangian-m as the service distribution of the CPU, we replace the CPU server by m identical servers. For example if we use the Erlangian-2 instead of the exponential distribution in the previous example, then our basic matrices are: CPU CPU2 Disk Comm R.disk p( p2( P =,

5 M µ cpu = µ cpu 2 µ D µ com µ RD p = [,,,, ], q = [, q,,, ] The value of t cpu is still the same as before and t cpu, t cpu2 are t cpu = t cpu2 = t cpu / 2 We can use the same approach to the local disk, remote disk or communication channel Hyperexponential Distributions The other class is the family of Hyperexponential distributions whose pdf is of the form f Hm (t) :=p µ exp(-µ t) + p 2 µ 2 exp(-µ 2 t) m p m µ m exp(-µ m t) = pi[ µ i exp( µ it )] i= where, p i and µ i are real and p +.p m =. Considering the same example but with the Hyperexponential-2(H 2 ) as the service distribution of the CPU. The basic matrices are: p = [p H2, -p H2,,, ] q = [q, q,,, ] P CPU CPU2 Disk Comm R.disk = ph2 ph2 µ cpu = M µ cpu -p -p 2 H2 H2 µ D p( p( µ com p2( p2( µ RD We need to calculate three parameters µ, µ 2 and p H2. Once E(T H2 ) = t cpu and the variance (σ 2 ) are chosen, one more parameter is still needed. One possibility is to fix the third parameter based on the physical system (i.e. specific value for p H2 ) or use the third moment. Another possibility is to fit the value of the pdf at. That is, let f H2 () = p H2 µ + ( - p H2 ) µ 2 Once the basic matrices are constructed, we can construct the other matrices that are to be used to represent a cluster of K servers (P k, M k, Q k, R k ) and K active tasks. See [7] for more details. Ξ k := { i = (α, α 2, α 3, α k ) α j k, K and α j = k } k K j= Ξ k is the set of all internal states of the system when there are k active customers there. Each m-tuple represents a state where α j is the number of tasks at server j. [M k ] ii = α µ + α 2 µ 2 + α 3 µ 3 + α k µ k = α j µ j j= [P k ] ii :=, unless [(i) (j)] has exactly two nonzero elements, one with the value and the other is. This means that only one task can move at a time. Let α a be the number of customers in the server where the task left and α b is the number of customers in the server where the task went. Then, [P k ] ij = [P] ab α aµ a [M ] k ii [R k ] ij =, unless [(j) (i)] has exactly one nonzero element and that element would have the value. Let a be the component that is not zero. [R k ] ij = p a [Q k ] ij =, unless [(i) (j)] has exactly one nonzero element and that element would have the value. Let a be the component that is not zero. [Q k ] ij = α aµ a q [ M ] a k ii 5. Reliability analysis The model presented in this paper can be used in reliability analysis of systems with failure without repair as follows: - First, Identify the system and application parameters. - Apply the model and get the initial expected running time for the target application. - Assume a failure distribution. - After the first epoch, calculate the failure probability of each server in the system. - Modify the entrance vector then recalculate the expected running time for the remaining tasks. - Repeat the previous two steps after each epoch. - Add more resources if the expected running time does not satisfy the application demand. A reliability model for simple systems with failures and repair has introduced in []. We are presently working on combining the two models. K

6 6. Results In this section, we show some of the results that can be obtained from the model. First, we show the different performance regions (transient, steady state and draining). Then, we demonstrate the effect of the distribution on the steady state value of the system. Finally, we study the effect of the service distribution and the performance region on some of the performance metrics (speedup and execution time prediction). We calculated two sets of examples one for K=5 and one for K=8. The average execution time per task is 2 units of time (E(T)=2). The application is running on a 5-node and 8-node cluster. To show the effect of the performance region on the accuracy of performance measures, we set N = 3 and. The results we show are for a central cluster. 6. Shared servers with non-exponential service times We assume that the shared server (remote disk) is the non-exponential server while the dedicated servers are exponential servers. In this case Jackson networks models can t be applied. 6.. Performance behavior In Figures,2, we show how the performance behavior of an application changes if we assume different service distributions. As mentioned earlier, there are three different regions in the performance characteristic of any system. At the beginning, the system is idle and a set of tasks start to execute. The system begins its transient region and stays there for some time. Exp H2,C2= H2,C2=5 execution queue. Once the number of tasks remaining in the execution queue becomes less than the degree of multitasking, the draining region starts. The draining continues until the system finishes all of the tasks. In all the cases that we show in this paper, the region where the interdeparture time is appears to be constant is the steady state region. Time 4 Exp H2,C2= H2,C2= Task Order Fig task application running on 8-node cluster. In Figures and 2, we compare exponential (C 2 = ), with Hyperexponential-2 with (C 2 = ) and (C 2 = 5). C 2 is the coefficient of variation. In all the cases we show, the time axis (the interdeparture time) is in log scale, the interdeparture time is the time between two successive task completions. Note that the steady state values in Figures and 2 did change that much even though we increased the K from 5 to 8 workstations. This happens because the contention at the central disk Performance prediction One of the main objectives of having a performance model is to be able to predict the running time of the target application. In most of the cases, the exponential distribution is assumed. As we mentioned earlier, it has been shown that the exponential assumption is not recommended for accurate estimations. Since parallel applications tend to have a wide range of execution times (large C 2 ). Time 5 K = 5 4 Task Order Fig. 3-task application running on 5-node cluster. After a number of tasks are processed the system approaches the steady state region. The system continues to execute tasks as they are available in the E% 3 2 N=3 N= C 2 Fig. 3. Prediction Error

7 Here, we give some examples to show the effect of assuming the exponential assumption where it is not valid. In Figures 3 and 4, we calculate the percentage error of the application, if the exponential distribution is assumed. The percentage error is calculated as : E = E E( Tact) ( Tact) app E( T exp app ) app *, where, E(T act ) app is the total execution time of the parallel application with the right distribution. E(T exp ) app is the total execution time of the parallel application if the exponential distribution is assumed. E% K = 8 N=3 N= C 2 Fig.4. Prediction Error Our results indicate that the exponential distribution fails to approximate distributions with large C 2. In Fig 3,4 we see that the error exceeds 2 % if C 2 =. The percentage error always increases with increasing C Speedup It always important to know the value of the speedup that your computing resources can provide. We found that one of the common problems is that the actual speedup is less than the expected speedup from the available resources. Our results show that there are three reasons. Obviously the first reason is contention. The second reason is the operating region (i.e., steady state or transient). In Figures 5 and 6, we use the same application but with different number of tasks (3, ) and calculate the speed up for different C 2. In case of 3 tasks, the transient region is dominating. Then we increased the number of tasks to in order for the steady state region to dominate. It is clear that, if the system is working in the transient region, the speed up is much less than if the system is working in the steady state region. This behavior can be explained if we return to Figures to 2 and notice the effect of the transient and draining regions explained earlier. Speedup Speedup K = C 2 Fig 5. Speedup K = 8 N=3 N= C 2 Fig. 6. Speedup N=3 N= The third reason for not getting the expected speed up is using the incorrect service distribution. It is clear that, the exponential distribution overestimates the speedup that we get if we use Hyperexponential distributions. This explains why we do not get the expected speedup from the system. We believe that both system operating region and the service distribution are equally important in analyzing the performance of parallel and distributed systems. 6.2 Dedicated Servers with Non-Exponential Service Times Here we assume that the dedicated servers (e.g. CPU) are the non-exponential server while the shared servers are exponential servers, reported in [5]. In this case Jackson networks models can be applied and our model is considered as a transient model for Jackson networks. Here, the steady state value of the system is the same for all distributions and equals the steady state value of the PF (steady state) solution. For results and analysis, see [5].

8 Conclusion In this paper, we presented an analytical performance model that can be used as a transient model for Jackson networks. Also, it can be used in cases where Jackson networks can t be applied. The model can be applied to systems with population size constraints. The model can use any service distribution and is not limited to exponential distributions. We also used the model to analyze the performance of parallel and distributed systems. Our results indicate that system operating region (steady state or transient) and service distribution can have a significant impact on the performance behavior of parallel and distributed systems. We showed how performance metrics like speedup and execution time prediction significantly depend on system operating region and service distribution. We then showed to what extent the exponential distribution can be used to approximate other service distributions. Finally, we utilized the model to be used in reliability analysis. REFERENCES []Ahmed Mohamed, Lester Lipsky and Reda Ammar, Performance Model for a Cluster of Workstations. The Fourth International Conference on Communications in Computing (CIC 23). Las Vegas, NV, Jun. 23. [2] Ahmed Mohamed, Lester Lipsky and Reda Ammar, Efficient Data Allocation for a Cluster of Workstations. ISCA 6th International Conference on Parallel and Distributed Computing Systems, Reno, NV, Aug 3. [3] Ahmed Mohamed, Lester Lipsky and Reda Ammar, Transient Model for Jackson Networks and its Application in Cluster Computing, Submitted to Journal of cluster Computing, Nov. 23. [4] Ahmed Mohamed, Lester Lipsky and Reda Ammar, Transient Model for Jackson Networks and its Approximation, 7 th International Conference on Principles of Distributed Systems (OPODIS3), Dec. 23. [5] Ahmed Mohamed, Lester Lipsky and Reda Ammar, Jackson Networks as Approximations to Parallel and Distributed Systems, Submitted to Sigmatrix, Nov. 23. [6] R. Buyya, High Performance Cluster Computing: Architecture and Systems, Prentice Hall PTR, NJ, 999. [7]J.Buzen, Queueing Network Models of Multiprogramming, Ph.D. Thesis, Harvard University, 97. [8] J. Buzen, Computational Algorithms for Closed Queueing, Comm. ACM, Vol 6, No. 9, Sep 973 [9] R. Chen, A Hybrid Solution of Fork/Join Synchronization in Parallel Queues, IEEE Transactions on Parallel and Distributed Systems, vol. 2, no. 8, pp , Aug. 2. [] P. Fiorini and C. Campbell and Lester Lipsky, An Analytic Approach to Assess the Performability of Parallel and Distributed Systems 5th International Conference on Parallel and Distributed Computing Systems, Louisville, Ky, Sep. 9-2, 22. [] I. Foster and Kesselman, The Grid: Blueprint for a New Computing Infrastructure, Morgan-Kaufmann, 998 [2] W. Gordon, G. Newell, Closed Queueing Systems, JORSA, Vol. 5, pp , 967. [3] J. Hatem and Lester Lipsky, Buffer Problems on Telecommunications Networks, 5 th International Conference on Telecommunication Systems, Nashville, TN, 997. [4] J. Jackson, Jopshop-Like Queueing Systems, J. TIMS, Vol., pp. 3-42, 963. [5] Leland, T. Ott, Analysis of CPU times on 6 VAX /78 at BELLCORE, Proceeding of the International Conference on Measurements and Modeling, April 986. [6] Lester Lipsyk, The Importance of Power-Tail Distributions for Modeling Queueing Systems,, Operations Research, Vol 47, No. 2, (March-April 999). [7] Lester Lipsky, "QUEUEING THEORY: A Linear Algebraic Approach", McMillan, New York, 992. [8] Lester Lipsky and J. Church, Applications of Queueing Network Models," Comp. Surveys, 9, pp , Sep [9] Lester Lipsky and Mark E. Crovela, "Long-Lasting Transient Conditions in Simulations with Heavy-Tailed Workloads," in Proceedings of the 997 Winter Simulation Conference, December 997. [2] F. Basket, R. Muntz, K. Chandy, Open, closed and Mixed Networks of Queues with Different Classes of Customers, JACM, Vol. 22, pp , Apr 975. [2] A. Tehranipour,, The Generalized M/G/C//N-Queue as a Model for Time-Sharing Systems, ACM-IEEE Joint Symposium on Applied Computing, Fayetteville, Apr, 99. [22] K. S. Trividi, Probability & Statistics with Reliability, Queueing and Computer Science Applications, Prentice-Hall, Inc., New Jersey, 982. [23] Qin, Sholl, Ammar, Micro Time Cost Analysis of Parallel Computations, IEEE Trans. Computers, vol. 4, no. 5, pp , May 99. [24] Yan, Zhang, Song, An Effective and Practical Performance Prediction Model for Parallel Computing on Nondedicated Heterogeneous Networks of Workstations, J. Para Dist Computing, vol.38, No., pp. 63-8, 996. [25] T. Zhang, S. Kang and Lester Lipsky,, On The Performance of Parallel Computers: Order Statistics and Amdahl s Law, International Journal Of Computers And Their Applications, Vol 3, No. 2, August 996.

Convolution Algorithm

Convolution Algorithm Convolution Algorithm Raj Jain Washington University in Saint Louis Saint Louis, MO 63130 Jain@cse.wustl.edu Audio/Video recordings of this lecture are available at: http://www.cse.wustl.edu/~jain/cse567-08/

More information

Analysis of Software Artifacts

Analysis of Software Artifacts Analysis of Software Artifacts System Performance I Shu-Ngai Yeung (with edits by Jeannette Wing) Department of Statistics Carnegie Mellon University Pittsburgh, PA 15213 2001 by Carnegie Mellon University

More information

Module 5: CPU Scheduling

Module 5: CPU Scheduling Module 5: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation 5.1 Basic Concepts Maximum CPU utilization obtained

More information

Glossary availability cellular manufacturing closed queueing network coefficient of variation (CV) conditional probability CONWIP

Glossary availability cellular manufacturing closed queueing network coefficient of variation (CV) conditional probability CONWIP Glossary availability The long-run average fraction of time that the processor is available for processing jobs, denoted by a (p. 113). cellular manufacturing The concept of organizing the factory into

More information

QUEUING MODELS AND MARKOV PROCESSES

QUEUING MODELS AND MARKOV PROCESSES QUEUING MODELS AND MARKOV ROCESSES Queues form when customer demand for a service cannot be met immediately. They occur because of fluctuations in demand levels so that models of queuing are intrinsically

More information

Queueing systems. Renato Lo Cigno. Simulation and Performance Evaluation Queueing systems - Renato Lo Cigno 1

Queueing systems. Renato Lo Cigno. Simulation and Performance Evaluation Queueing systems - Renato Lo Cigno 1 Queueing systems Renato Lo Cigno Simulation and Performance Evaluation 2014-15 Queueing systems - Renato Lo Cigno 1 Queues A Birth-Death process is well modeled by a queue Indeed queues can be used to

More information

Operational Laws Raj Jain

Operational Laws Raj Jain Operational Laws 33-1 Overview What is an Operational Law? 1. Utilization Law 2. Forced Flow Law 3. Little s Law 4. General Response Time Law 5. Interactive Response Time Law 6. Bottleneck Analysis 33-2

More information

COMP9334: Capacity Planning of Computer Systems and Networks

COMP9334: Capacity Planning of Computer Systems and Networks COMP9334: Capacity Planning of Computer Systems and Networks Week 2: Operational analysis Lecturer: Prof. Sanjay Jha NETWORKS RESEARCH GROUP, CSE, UNSW Operational analysis Operational: Collect performance

More information

How to deal with uncertainties and dynamicity?

How to deal with uncertainties and dynamicity? How to deal with uncertainties and dynamicity? http://graal.ens-lyon.fr/ lmarchal/scheduling/ 19 novembre 2012 1/ 37 Outline 1 Sensitivity and Robustness 2 Analyzing the sensitivity : the case of Backfilling

More information

Chapter 6: CPU Scheduling

Chapter 6: CPU Scheduling Chapter 6: CPU Scheduling Basic Concepts Scheduling Criteria Scheduling Algorithms Multiple-Processor Scheduling Real-Time Scheduling Algorithm Evaluation 6.1 Basic Concepts Maximum CPU utilization obtained

More information

Queues and Queueing Networks

Queues and Queueing Networks Queues and Queueing Networks Sanjay K. Bose Dept. of EEE, IITG Copyright 2015, Sanjay K. Bose 1 Introduction to Queueing Models and Queueing Analysis Copyright 2015, Sanjay K. Bose 2 Model of a Queue Arrivals

More information

CSM: Operational Analysis

CSM: Operational Analysis CSM: Operational Analysis 2016-17 Computer Science Tripos Part II Computer Systems Modelling: Operational Analysis by Ian Leslie Richard Gibbens, Ian Leslie Operational Analysis Based on the idea of observation

More information

CPU scheduling. CPU Scheduling

CPU scheduling. CPU Scheduling EECS 3221 Operating System Fundamentals No.4 CPU scheduling Prof. Hui Jiang Dept of Electrical Engineering and Computer Science, York University CPU Scheduling CPU scheduling is the basis of multiprogramming

More information

Exercises Solutions. Automation IEA, LTH. Chapter 2 Manufacturing and process systems. Chapter 5 Discrete manufacturing problems

Exercises Solutions. Automation IEA, LTH. Chapter 2 Manufacturing and process systems. Chapter 5 Discrete manufacturing problems Exercises Solutions Note, that we have not formulated the answers for all the review questions. You will find the answers for many questions by reading and reflecting about the text in the book. Chapter

More information

Che-Wei Chang Department of Computer Science and Information Engineering, Chang Gung University

Che-Wei Chang Department of Computer Science and Information Engineering, Chang Gung University Che-Wei Chang chewei@mail.cgu.edu.tw Department of Computer Science and Information Engineering, Chang Gung University } 2017/11/15 Midterm } 2017/11/22 Final Project Announcement 2 1. Introduction 2.

More information

CPSC 531: System Modeling and Simulation. Carey Williamson Department of Computer Science University of Calgary Fall 2017

CPSC 531: System Modeling and Simulation. Carey Williamson Department of Computer Science University of Calgary Fall 2017 CPSC 531: System Modeling and Simulation Carey Williamson Department of Computer Science University of Calgary Fall 2017 Motivating Quote for Queueing Models Good things come to those who wait - poet/writer

More information

6 Solving Queueing Models

6 Solving Queueing Models 6 Solving Queueing Models 6.1 Introduction In this note we look at the solution of systems of queues, starting with simple isolated queues. The benefits of using predefined, easily classified queues will

More information

Queueing Theory I Summary! Little s Law! Queueing System Notation! Stationary Analysis of Elementary Queueing Systems " M/M/1 " M/M/m " M/M/1/K "

Queueing Theory I Summary! Little s Law! Queueing System Notation! Stationary Analysis of Elementary Queueing Systems  M/M/1  M/M/m  M/M/1/K Queueing Theory I Summary Little s Law Queueing System Notation Stationary Analysis of Elementary Queueing Systems " M/M/1 " M/M/m " M/M/1/K " Little s Law a(t): the process that counts the number of arrivals

More information

Andrew Morton University of Waterloo Canada

Andrew Morton University of Waterloo Canada EDF Feasibility and Hardware Accelerators Andrew Morton University of Waterloo Canada Outline 1) Introduction and motivation 2) Review of EDF and feasibility analysis 3) Hardware accelerators and scheduling

More information

A PARAMETRIC DECOMPOSITION BASED APPROACH FOR MULTI-CLASS CLOSED QUEUING NETWORKS WITH SYNCHRONIZATION STATIONS

A PARAMETRIC DECOMPOSITION BASED APPROACH FOR MULTI-CLASS CLOSED QUEUING NETWORKS WITH SYNCHRONIZATION STATIONS A PARAMETRIC DECOMPOSITION BASED APPROACH FOR MULTI-CLASS CLOSED QUEUING NETWORKS WITH SYNCHRONIZATION STATIONS Kumar Satyam and Ananth Krishnamurthy Department of Decision Sciences and Engineering Systems,

More information

PERFORMANCE METRICS. Mahdi Nazm Bojnordi. CS/ECE 6810: Computer Architecture. Assistant Professor School of Computing University of Utah

PERFORMANCE METRICS. Mahdi Nazm Bojnordi. CS/ECE 6810: Computer Architecture. Assistant Professor School of Computing University of Utah PERFORMANCE METRICS Mahdi Nazm Bojnordi Assistant Professor School of Computing University of Utah CS/ECE 6810: Computer Architecture Overview Announcement Jan. 17 th : Homework 1 release (due on Jan.

More information

COMP9334 Capacity Planning for Computer Systems and Networks

COMP9334 Capacity Planning for Computer Systems and Networks COMP9334 Capacity Planning for Computer Systems and Networks Week 2: Operational Analysis and Workload Characterisation COMP9334 1 Last lecture Modelling of computer systems using Queueing Networks Open

More information

Intro Refresher Reversibility Open networks Closed networks Multiclass networks Other networks. Queuing Networks. Florence Perronnin

Intro Refresher Reversibility Open networks Closed networks Multiclass networks Other networks. Queuing Networks. Florence Perronnin Queuing Networks Florence Perronnin Polytech Grenoble - UGA March 23, 27 F. Perronnin (UGA) Queuing Networks March 23, 27 / 46 Outline Introduction to Queuing Networks 2 Refresher: M/M/ queue 3 Reversibility

More information

EE 368. Weeks 3 (Notes)

EE 368. Weeks 3 (Notes) EE 368 Weeks 3 (Notes) 1 State of a Queuing System State: Set of parameters that describe the condition of the system at a point in time. Why do we need it? Average size of Queue Average waiting time How

More information

Scheduling I. Today. Next Time. ! Introduction to scheduling! Classical algorithms. ! Advanced topics on scheduling

Scheduling I. Today. Next Time. ! Introduction to scheduling! Classical algorithms. ! Advanced topics on scheduling Scheduling I Today! Introduction to scheduling! Classical algorithms Next Time! Advanced topics on scheduling Scheduling out there! You are the manager of a supermarket (ok, things don t always turn out

More information

Slides 9: Queuing Models

Slides 9: Queuing Models Slides 9: Queuing Models Purpose Simulation is often used in the analysis of queuing models. A simple but typical queuing model is: Queuing models provide the analyst with a powerful tool for designing

More information

EP2200 Course Project 2017 Project II - Mobile Computation Offloading

EP2200 Course Project 2017 Project II - Mobile Computation Offloading EP2200 Course Project 2017 Project II - Mobile Computation Offloading 1 Introduction Queuing theory provides us a very useful mathematic tool that can be used to analytically evaluate the performance of

More information

57:022 Principles of Design II Final Exam Solutions - Spring 1997

57:022 Principles of Design II Final Exam Solutions - Spring 1997 57:022 Principles of Design II Final Exam Solutions - Spring 1997 Part: I II III IV V VI Total Possible Pts: 52 10 12 16 13 12 115 PART ONE Indicate "+" if True and "o" if False: + a. If a component's

More information

Performance Evaluation of Queuing Systems

Performance Evaluation of Queuing Systems Performance Evaluation of Queuing Systems Introduction to Queuing Systems System Performance Measures & Little s Law Equilibrium Solution of Birth-Death Processes Analysis of Single-Station Queuing Systems

More information

On Stability and Sojourn Time of Peer-to-Peer Queuing Systems

On Stability and Sojourn Time of Peer-to-Peer Queuing Systems On Stability and Sojourn Time of Peer-to-Peer Queuing Systems Taoyu Li Minghua Chen Tony Lee Xing Li Tsinghua University, Beijing, China. {ldy03@mails.tsinghua.edu.cn,xing@cernet.edu.cn} The Chinese University

More information

Chapter 6 Queueing Models. Banks, Carson, Nelson & Nicol Discrete-Event System Simulation

Chapter 6 Queueing Models. Banks, Carson, Nelson & Nicol Discrete-Event System Simulation Chapter 6 Queueing Models Banks, Carson, Nelson & Nicol Discrete-Event System Simulation Purpose Simulation is often used in the analysis of queueing models. A simple but typical queueing model: Queueing

More information

Scheduling I. Today Introduction to scheduling Classical algorithms. Next Time Advanced topics on scheduling

Scheduling I. Today Introduction to scheduling Classical algorithms. Next Time Advanced topics on scheduling Scheduling I Today Introduction to scheduling Classical algorithms Next Time Advanced topics on scheduling Scheduling out there You are the manager of a supermarket (ok, things don t always turn out the

More information

Buzen s algorithm. Cyclic network Extension of Jackson networks

Buzen s algorithm. Cyclic network Extension of Jackson networks Outline Buzen s algorithm Mean value analysis for Jackson networks Cyclic network Extension of Jackson networks BCMP network 1 Marginal Distributions based on Buzen s algorithm With Buzen s algorithm,

More information

Multi-State Availability Modeling in Practice

Multi-State Availability Modeling in Practice Multi-State Availability Modeling in Practice Kishor S. Trivedi, Dong Seong Kim, Xiaoyan Yin Depart ment of Electrical and Computer Engineering, Duke University, Durham, NC 27708 USA kst@ee.duke.edu, {dk76,

More information

Notation. Bounds on Speedup. Parallel Processing. CS575 Parallel Processing

Notation. Bounds on Speedup. Parallel Processing. CS575 Parallel Processing Parallel Processing CS575 Parallel Processing Lecture five: Efficiency Wim Bohm, Colorado State University Some material from Speedup vs Efficiency in Parallel Systems - Eager, Zahorjan and Lazowska IEEE

More information

CPSC 531 Systems Modeling and Simulation FINAL EXAM

CPSC 531 Systems Modeling and Simulation FINAL EXAM CPSC 531 Systems Modeling and Simulation FINAL EXAM Department of Computer Science University of Calgary Professor: Carey Williamson December 21, 2017 This is a CLOSED BOOK exam. Textbooks, notes, laptops,

More information

CHAPTER 3 STOCHASTIC MODEL OF A GENERAL FEED BACK QUEUE NETWORK

CHAPTER 3 STOCHASTIC MODEL OF A GENERAL FEED BACK QUEUE NETWORK CHAPTER 3 STOCHASTIC MODEL OF A GENERAL FEED BACK QUEUE NETWORK 3. INTRODUCTION: Considerable work has been turned out by Mathematicians and operation researchers in the development of stochastic and simulation

More information

Stochastic process. X, a series of random variables indexed by t

Stochastic process. X, a series of random variables indexed by t Stochastic process X, a series of random variables indexed by t X={X(t), t 0} is a continuous time stochastic process X={X(t), t=0,1, } is a discrete time stochastic process X(t) is the state at time t,

More information

Process-Algebraic Modelling of Priority Queueing Networks

Process-Algebraic Modelling of Priority Queueing Networks Process-Algebraic Modelling of Priority Queueing Networks Giuliano Casale Department of Computing Imperial College ondon, U.K. gcasale@doc.ic.ac.uk Mirco Tribastone Institut für Informatik udwig-maximilians-universität

More information

THE HEAVY-TRAFFIC BOTTLENECK PHENOMENON IN OPEN QUEUEING NETWORKS. S. Suresh and W. Whitt AT&T Bell Laboratories Murray Hill, New Jersey 07974

THE HEAVY-TRAFFIC BOTTLENECK PHENOMENON IN OPEN QUEUEING NETWORKS. S. Suresh and W. Whitt AT&T Bell Laboratories Murray Hill, New Jersey 07974 THE HEAVY-TRAFFIC BOTTLENECK PHENOMENON IN OPEN QUEUEING NETWORKS by S. Suresh and W. Whitt AT&T Bell Laboratories Murray Hill, New Jersey 07974 ABSTRACT This note describes a simulation experiment involving

More information

Recap. Probability, stochastic processes, Markov chains. ELEC-C7210 Modeling and analysis of communication networks

Recap. Probability, stochastic processes, Markov chains. ELEC-C7210 Modeling and analysis of communication networks Recap Probability, stochastic processes, Markov chains ELEC-C7210 Modeling and analysis of communication networks 1 Recap: Probability theory important distributions Discrete distributions Geometric distribution

More information

In Proceedings of the 1997 Winter Simulation Conference, S. Andradottir, K. J. Healy, D. H. Withers, and B. L. Nelson, eds.

In Proceedings of the 1997 Winter Simulation Conference, S. Andradottir, K. J. Healy, D. H. Withers, and B. L. Nelson, eds. In Proceedings of the 1997 Winter Simulation Conference, S. Andradottir, K. J. Healy, D. H. Withers, and B. L. Nelson, eds. LONG-LASTING TRANSIENT CONDITIONS IN SIMULATIONS WITH HEAVY-TAILED WORKLOADS

More information

ON THE NON-EXISTENCE OF PRODUCT-FORM SOLUTIONS FOR QUEUEING NETWORKS WITH RETRIALS

ON THE NON-EXISTENCE OF PRODUCT-FORM SOLUTIONS FOR QUEUEING NETWORKS WITH RETRIALS ON THE NON-EXISTENCE OF PRODUCT-FORM SOLUTIONS FOR QUEUEING NETWORKS WITH RETRIALS J.R. ARTALEJO, Department of Statistics and Operations Research, Faculty of Mathematics, Complutense University of Madrid,

More information

Computer Systems Modelling

Computer Systems Modelling Computer Systems Modelling Computer Laboratory Computer Science Tripos, Part II Michaelmas Term 2003 R. J. Gibbens Problem sheet William Gates Building JJ Thomson Avenue Cambridge CB3 0FD http://www.cl.cam.ac.uk/

More information

Section 3.3: Discrete-Event Simulation Examples

Section 3.3: Discrete-Event Simulation Examples Section 33: Discrete-Event Simulation Examples Discrete-Event Simulation: A First Course c 2006 Pearson Ed, Inc 0-13-142917-5 Discrete-Event Simulation: A First Course Section 33: Discrete-Event Simulation

More information

Tuan V. Dinh, Lachlan Andrew and Philip Branch

Tuan V. Dinh, Lachlan Andrew and Philip Branch Predicting supercomputing workload using per user information Tuan V. Dinh, Lachlan Andrew and Philip Branch 13 th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing. Delft, 14 th -16

More information

A POPULATION-MIX DRIVEN APPROXIMATION FOR QUEUEING NETWORKS WITH FINITE CAPACITY REGIONS

A POPULATION-MIX DRIVEN APPROXIMATION FOR QUEUEING NETWORKS WITH FINITE CAPACITY REGIONS A POPULATION-MIX DRIVEN APPROXIMATION FOR QUEUEING NETWORKS WITH FINITE CAPACITY REGIONS J. Anselmi 1, G. Casale 2, P. Cremonesi 1 1 Politecnico di Milano, Via Ponzio 34/5, I-20133 Milan, Italy 2 Neptuny

More information

Quality of Real-Time Streaming in Wireless Cellular Networks : Stochastic Modeling and Analysis

Quality of Real-Time Streaming in Wireless Cellular Networks : Stochastic Modeling and Analysis Quality of Real-Time Streaming in Wireless Cellular Networs : Stochastic Modeling and Analysis B. Blaszczyszyn, M. Jovanovic and M. K. Karray Based on paper [1] WiOpt/WiVid Mai 16th, 2014 Outline Introduction

More information

Journal of Global Research in Computer Science

Journal of Global Research in Computer Science Volume 2, No. 4, April 2011 Journal of Global Research in Computer Science RESEARCH ARTICLE Available Online at www.jgrcs.info PERFORMANCE EVALUATION OF A NEW PROPOSED SHOTREST EXECUTION FIRST DYNAMIC

More information

An Analysis of the Preemptive Repeat Queueing Discipline

An Analysis of the Preemptive Repeat Queueing Discipline An Analysis of the Preemptive Repeat Queueing Discipline Tony Field August 3, 26 Abstract An analysis of two variants of preemptive repeat or preemptive repeat queueing discipline is presented: one in

More information

IEOR 4106: Spring Solutions to Homework Assignment 7: Due on Tuesday, March 22.

IEOR 4106: Spring Solutions to Homework Assignment 7: Due on Tuesday, March 22. IEOR 46: Spring Solutions to Homework Assignment 7: Due on Tuesday, March. More of Chapter 5: Read the rest of Section 5.3, skipping Examples 5.7 (Coupon Collecting), 5. (Insurance claims)and Subsection

More information

Introduction to Markov Chains, Queuing Theory, and Network Performance

Introduction to Markov Chains, Queuing Theory, and Network Performance Introduction to Markov Chains, Queuing Theory, and Network Performance Marceau Coupechoux Telecom ParisTech, departement Informatique et Réseaux marceau.coupechoux@telecom-paristech.fr IT.2403 Modélisation

More information

CPU SCHEDULING RONG ZHENG

CPU SCHEDULING RONG ZHENG CPU SCHEDULING RONG ZHENG OVERVIEW Why scheduling? Non-preemptive vs Preemptive policies FCFS, SJF, Round robin, multilevel queues with feedback, guaranteed scheduling 2 SHORT-TERM, MID-TERM, LONG- TERM

More information

Availability. M(t) = 1 - e -mt

Availability. M(t) = 1 - e -mt Availability Availability - A(t) the probability that the system is operating correctly and is available to perform its functions at the instant of time t More general concept than reliability: failure

More information

CPU Scheduling. CPU Scheduler

CPU Scheduling. CPU Scheduler CPU Scheduling These slides are created by Dr. Huang of George Mason University. Students registered in Dr. Huang s courses at GMU can make a single machine readable copy and print a single copy of each

More information

Queuing Networks: Burke s Theorem, Kleinrock s Approximation, and Jackson s Theorem. Wade Trappe

Queuing Networks: Burke s Theorem, Kleinrock s Approximation, and Jackson s Theorem. Wade Trappe Queuing Networks: Burke s Theorem, Kleinrock s Approximation, and Jackson s Theorem Wade Trappe Lecture Overview Network of Queues Introduction Queues in Tandem roduct Form Solutions Burke s Theorem What

More information

EEC 686/785 Modeling & Performance Evaluation of Computer Systems. Lecture 19

EEC 686/785 Modeling & Performance Evaluation of Computer Systems. Lecture 19 EEC 686/785 Modeling & Performance Evaluation of Computer Systems Lecture 19 Department of Electrical and Computer Engineering Cleveland State University wenbing@ieee.org (based on Dr. Raj Jain s lecture

More information

Load Balancing in Distributed Service System: A Survey

Load Balancing in Distributed Service System: A Survey Load Balancing in Distributed Service System: A Survey Xingyu Zhou The Ohio State University zhou.2055@osu.edu November 21, 2016 Xingyu Zhou (OSU) Load Balancing November 21, 2016 1 / 29 Introduction and

More information

Energy-efficient Mapping of Big Data Workflows under Deadline Constraints

Energy-efficient Mapping of Big Data Workflows under Deadline Constraints Energy-efficient Mapping of Big Data Workflows under Deadline Constraints Presenter: Tong Shu Authors: Tong Shu and Prof. Chase Q. Wu Big Data Center Department of Computer Science New Jersey Institute

More information

Process Scheduling. Process Scheduling. CPU and I/O Bursts. CPU - I/O Burst Cycle. Variations in Bursts. Histogram of CPU Burst Times

Process Scheduling. Process Scheduling. CPU and I/O Bursts. CPU - I/O Burst Cycle. Variations in Bursts. Histogram of CPU Burst Times Scheduling The objective of multiprogramming is to have some process running all the time The objective of timesharing is to have the switch between processes so frequently that users can interact with

More information

The exponential distribution and the Poisson process

The exponential distribution and the Poisson process The exponential distribution and the Poisson process 1-1 Exponential Distribution: Basic Facts PDF f(t) = { λe λt, t 0 0, t < 0 CDF Pr{T t) = 0 t λe λu du = 1 e λt (t 0) Mean E[T] = 1 λ Variance Var[T]

More information

Last class: Today: Threads. CPU Scheduling

Last class: Today: Threads. CPU Scheduling 1 Last class: Threads Today: CPU Scheduling 2 Resource Allocation In a multiprogramming system, we need to share resources among the running processes What are the types of OS resources? Question: Which

More information

Part I Stochastic variables and Markov chains

Part I Stochastic variables and Markov chains Part I Stochastic variables and Markov chains Random variables describe the behaviour of a phenomenon independent of any specific sample space Distribution function (cdf, cumulative distribution function)

More information

HITTING TIME IN AN ERLANG LOSS SYSTEM

HITTING TIME IN AN ERLANG LOSS SYSTEM Probability in the Engineering and Informational Sciences, 16, 2002, 167 184+ Printed in the U+S+A+ HITTING TIME IN AN ERLANG LOSS SYSTEM SHELDON M. ROSS Department of Industrial Engineering and Operations

More information

Outline. Simulation of a Single-Server Queueing System. EEC 686/785 Modeling & Performance Evaluation of Computer Systems.

Outline. Simulation of a Single-Server Queueing System. EEC 686/785 Modeling & Performance Evaluation of Computer Systems. EEC 686/785 Modeling & Performance Evaluation of Computer Systems Lecture 19 Outline Simulation of a Single-Server Queueing System Review of midterm # Department of Electrical and Computer Engineering

More information

E-Companion to Fully Sequential Procedures for Large-Scale Ranking-and-Selection Problems in Parallel Computing Environments

E-Companion to Fully Sequential Procedures for Large-Scale Ranking-and-Selection Problems in Parallel Computing Environments E-Companion to Fully Sequential Procedures for Large-Scale Ranking-and-Selection Problems in Parallel Computing Environments Jun Luo Antai College of Economics and Management Shanghai Jiao Tong University

More information

An Algorithm for a Two-Disk Fault-Tolerant Array with (Prime 1) Disks

An Algorithm for a Two-Disk Fault-Tolerant Array with (Prime 1) Disks An Algorithm for a Two-Disk Fault-Tolerant Array with (Prime 1) Disks Sanjeeb Nanda and Narsingh Deo School of Computer Science University of Central Florida Orlando, Florida 32816-2362 sanjeeb@earthlink.net,

More information

Efficient Nonlinear Optimizations of Queuing Systems

Efficient Nonlinear Optimizations of Queuing Systems Efficient Nonlinear Optimizations of Queuing Systems Mung Chiang, Arak Sutivong, and Stephen Boyd Electrical Engineering Department, Stanford University, CA 9435 Abstract We present a systematic treatment

More information

A Study on Performance Analysis of Queuing System with Multiple Heterogeneous Servers

A Study on Performance Analysis of Queuing System with Multiple Heterogeneous Servers UNIVERSITY OF OKLAHOMA GENERAL EXAM REPORT A Study on Performance Analysis of Queuing System with Multiple Heterogeneous Servers Prepared by HUSNU SANER NARMAN husnu@ou.edu based on the papers 1) F. S.

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science MASSACHUSETTS INSTITUTE OF TECHNOLOGY Department of Electrical Engineering and Computer Science 6.262 Discrete Stochastic Processes Midterm Quiz April 6, 2010 There are 5 questions, each with several parts.

More information

Outline. Finite source queue M/M/c//K Queues with impatience (balking, reneging, jockeying, retrial) Transient behavior Advanced Queue.

Outline. Finite source queue M/M/c//K Queues with impatience (balking, reneging, jockeying, retrial) Transient behavior Advanced Queue. Outline Finite source queue M/M/c//K Queues with impatience (balking, reneging, jockeying, retrial) Transient behavior Advanced Queue Batch queue Bulk input queue M [X] /M/1 Bulk service queue M/M [Y]

More information

EDF Feasibility and Hardware Accelerators

EDF Feasibility and Hardware Accelerators EDF Feasibility and Hardware Accelerators Andrew Morton University of Waterloo, Waterloo, Canada, arrmorton@uwaterloo.ca Wayne M. Loucks University of Waterloo, Waterloo, Canada, wmloucks@pads.uwaterloo.ca

More information

Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing

Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing IEEE TRANSACTIONS ON SUSTAINABLE COMPUTING VOL XX NO YY MONTH 019 1 Computation Offloading Strategy Optimization with Multiple Heterogeneous Servers in Mobile Edge Computing Keqin Li Fellow IEEE Abstract

More information

Operational Laws 33-1

Operational Laws 33-1 Operational Laws Raj Jain Washington University in Saint Louis Jain@eecs.berkeley.edu or Jain@wustl.edu A Mini-Course offered at UC Berkeley, Sept-Oct 2012 These slides and audio/video recordings are available

More information

Welcome to MCS 572. content and organization expectations of the course. definition and classification

Welcome to MCS 572. content and organization expectations of the course. definition and classification Welcome to MCS 572 1 About the Course content and organization expectations of the course 2 Supercomputing definition and classification 3 Measuring Performance speedup and efficiency Amdahl s Law Gustafson

More information

Lecture Notes of Communication Network I y: part 5. Lijun Qian. Rutgers The State University of New Jersey. 94 Brett Road. Piscataway, NJ

Lecture Notes of Communication Network I y: part 5. Lijun Qian. Rutgers The State University of New Jersey. 94 Brett Road. Piscataway, NJ Lecture Notes of Communication Network I y: part 5 Lijun Qian Department of Electrical Engineering Rutgers The State University of New Jersey 94 Brett Road Piscataway, NJ 08854-8058 October 6, 1998 Abstract

More information

An M/M/1 Queue in Random Environment with Disasters

An M/M/1 Queue in Random Environment with Disasters An M/M/1 Queue in Random Environment with Disasters Noam Paz 1 and Uri Yechiali 1,2 1 Department of Statistics and Operations Research School of Mathematical Sciences Tel Aviv University, Tel Aviv 69978,

More information

Nikolai K. Krivulin. St.Petersburg State University

Nikolai K. Krivulin. St.Petersburg State University A MAX-ALGEBRA APPROACH TO MODELING AND SIMULATION OF TANDEM QUEUEING SYSTEMS Nikolai K Krivulin StPetersburg State University Abstract Max-algebra models of tandem single-server queueing systems with both

More information

Uniform random numbers generators

Uniform random numbers generators Uniform random numbers generators Lecturer: Dmitri A. Moltchanov E-mail: moltchan@cs.tut.fi http://www.cs.tut.fi/kurssit/tlt-2707/ OUTLINE: The need for random numbers; Basic steps in generation; Uniformly

More information

Analysis of Round-Robin Implementations of Processor Sharing, Including Overhead

Analysis of Round-Robin Implementations of Processor Sharing, Including Overhead Analysis of Round-Robin Implementations of Processor Sharing, Including Overhead Steve Thompson and Lester Lipsky Computer Science & Engineering University of Connecticut Storrs, CT 669 Email: sat3@engr.uconn.edu;

More information

Chapter 10. Queuing Systems. D (Queuing Theory) Queuing theory is the branch of operations research concerned with waiting lines.

Chapter 10. Queuing Systems. D (Queuing Theory) Queuing theory is the branch of operations research concerned with waiting lines. Chapter 10 Queuing Systems D. 10. 1. (Queuing Theory) Queuing theory is the branch of operations research concerned with waiting lines. D. 10.. (Queuing System) A ueuing system consists of 1. a user source.

More information

A Queueing System with Queue Length Dependent Service Times, with Applications to Cell Discarding in ATM Networks

A Queueing System with Queue Length Dependent Service Times, with Applications to Cell Discarding in ATM Networks A Queueing System with Queue Length Dependent Service Times, with Applications to Cell Discarding in ATM Networks by Doo Il Choi, Charles Knessl and Charles Tier University of Illinois at Chicago 85 South

More information

Advanced Computer Networks Lecture 3. Models of Queuing

Advanced Computer Networks Lecture 3. Models of Queuing Advanced Computer Networks Lecture 3. Models of Queuing Husheng Li Min Kao Department of Electrical Engineering and Computer Science University of Tennessee, Knoxville Spring, 2016 1/13 Terminology of

More information

Real-time operating systems course. 6 Definitions Non real-time scheduling algorithms Real-time scheduling algorithm

Real-time operating systems course. 6 Definitions Non real-time scheduling algorithms Real-time scheduling algorithm Real-time operating systems course 6 Definitions Non real-time scheduling algorithms Real-time scheduling algorithm Definitions Scheduling Scheduling is the activity of selecting which process/thread should

More information

Name of the Student:

Name of the Student: SUBJECT NAME : Probability & Queueing Theory SUBJECT CODE : MA 6453 MATERIAL NAME : Part A questions REGULATION : R2013 UPDATED ON : November 2017 (Upto N/D 2017 QP) (Scan the above QR code for the direct

More information

A Starvation-free Algorithm For Achieving 100% Throughput in an Input- Queued Switch

A Starvation-free Algorithm For Achieving 100% Throughput in an Input- Queued Switch A Starvation-free Algorithm For Achieving 00% Throughput in an Input- Queued Switch Abstract Adisak ekkittikul ick ckeown Department of Electrical Engineering Stanford University Stanford CA 9405-400 Tel

More information

Queuing Analysis of Markovian Queue Having Two Heterogeneous Servers with Catastrophes using Matrix Geometric Technique

Queuing Analysis of Markovian Queue Having Two Heterogeneous Servers with Catastrophes using Matrix Geometric Technique International Journal of Statistics and Systems ISSN 0973-2675 Volume 12, Number 2 (2017), pp. 205-212 Research India Publications http://www.ripublication.com Queuing Analysis of Markovian Queue Having

More information

Exercises Stochastic Performance Modelling. Hamilton Institute, Summer 2010

Exercises Stochastic Performance Modelling. Hamilton Institute, Summer 2010 Exercises Stochastic Performance Modelling Hamilton Institute, Summer Instruction Exercise Let X be a non-negative random variable with E[X ]

More information

Master thesis. Multi-class Fork-Join queues & The stochastic knapsack problem

Master thesis. Multi-class Fork-Join queues & The stochastic knapsack problem Master thesis Multi-class Fork-Join queues & The stochastic knapsack problem Sihan Ding August 26th, 2011 Supervisor UL: Dr. Floske Spieksma Supervisors CWI: Drs. Chrétien Verhoef Prof.dr. Rob van der

More information

One important issue in the study of queueing systems is to characterize departure processes. Study on departure processes was rst initiated by Burke (

One important issue in the study of queueing systems is to characterize departure processes. Study on departure processes was rst initiated by Burke ( The Departure Process of the GI/G/ Queue and Its MacLaurin Series Jian-Qiang Hu Department of Manufacturing Engineering Boston University 5 St. Mary's Street Brookline, MA 2446 Email: hqiang@bu.edu June

More information

A STAFFING ALGORITHM FOR CALL CENTERS WITH SKILL-BASED ROUTING: SUPPLEMENTARY MATERIAL

A STAFFING ALGORITHM FOR CALL CENTERS WITH SKILL-BASED ROUTING: SUPPLEMENTARY MATERIAL A STAFFING ALGORITHM FOR CALL CENTERS WITH SKILL-BASED ROUTING: SUPPLEMENTARY MATERIAL by Rodney B. Wallace IBM and The George Washington University rodney.wallace@us.ibm.com Ward Whitt Columbia University

More information

Data analysis and stochastic modeling

Data analysis and stochastic modeling Data analysis and stochastic modeling Lecture 7 An introduction to queueing theory Guillaume Gravier guillaume.gravier@irisa.fr with a lot of help from Paul Jensen s course http://www.me.utexas.edu/ jensen/ormm/instruction/powerpoint/or_models_09/14_queuing.ppt

More information

A FAST MATRIX-ANALYTIC APPROXIMATION FOR THE TWO CLASS GI/G/1 NON-PREEMPTIVE PRIORITY QUEUE

A FAST MATRIX-ANALYTIC APPROXIMATION FOR THE TWO CLASS GI/G/1 NON-PREEMPTIVE PRIORITY QUEUE A FAST MATRIX-ANAYTIC APPROXIMATION FOR TE TWO CASS GI/G/ NON-PREEMPTIVE PRIORITY QUEUE Gábor orváth Department of Telecommunication Budapest University of Technology and Economics. Budapest Pf. 9., ungary

More information

reversed chain is ergodic and has the same equilibrium probabilities (check that π j =

reversed chain is ergodic and has the same equilibrium probabilities (check that π j = Lecture 10 Networks of queues In this lecture we shall finally get around to consider what happens when queues are part of networks (which, after all, is the topic of the course). Firstly we shall need

More information

BIRTH DEATH PROCESSES AND QUEUEING SYSTEMS

BIRTH DEATH PROCESSES AND QUEUEING SYSTEMS BIRTH DEATH PROCESSES AND QUEUEING SYSTEMS Andrea Bobbio Anno Accademico 999-2000 Queueing Systems 2 Notation for Queueing Systems /λ mean time between arrivals S = /µ ρ = λ/µ N mean service time traffic

More information

Transient Solution of a Multi-Server Queue. with Catastrophes and Impatient Customers. when System is Down

Transient Solution of a Multi-Server Queue. with Catastrophes and Impatient Customers. when System is Down Applied Mathematical Sciences, Vol. 8, 2014, no. 92, 4585-4592 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2014.45388 Transient Solution of a Multi-Server Queue with Catastrophes and Impatient

More information

Exact Simulation of the Stationary Distribution of M/G/c Queues

Exact Simulation of the Stationary Distribution of M/G/c Queues 1/36 Exact Simulation of the Stationary Distribution of M/G/c Queues Professor Karl Sigman Columbia University New York City USA Conference in Honor of Søren Asmussen Monday, August 1, 2011 Sandbjerg Estate

More information

Push and Pull Systems in a Dynamic Environment

Push and Pull Systems in a Dynamic Environment Push and Pull Systems in a Dynamic Environment ichael Zazanis Dept. of IEOR University of assachusetts Amherst, A 0003 email: zazanis@ecs.umass.edu Abstract We examine Push and Pull production control

More information

The Expected Opportunity Cost and Selecting the Optimal Subset

The Expected Opportunity Cost and Selecting the Optimal Subset Applied Mathematical Sciences, Vol. 9, 2015, no. 131, 6507-6519 HIKARI Ltd, www.m-hikari.com http://dx.doi.org/10.12988/ams.2015.58561 The Expected Opportunity Cost and Selecting the Optimal Subset Mohammad

More information

SYMBOLS AND ABBREVIATIONS

SYMBOLS AND ABBREVIATIONS APPENDIX A SYMBOLS AND ABBREVIATIONS This appendix contains definitions of common symbols and abbreviations used frequently and consistently throughout the text. Symbols that are used only occasionally

More information