Optimal state-free, size-aware dispatching for heterogeneous -type systems

Feng, Huijie; Misra, Vishal; Rubenstein, Dan

doi:10.1016/j.peva.2005.07.031

Cited by 76 publications

(73 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…which gives (11). The first two terms in (11) correspond to the marginal cost of i 0 , the third term depends only on the first action i 0 but not on α 1 , and the last summation depends on both.…”

Section: Propositionmentioning

confidence: 99%

“…The first two terms in (11) correspond to the marginal cost of i 0 , the third term depends only on the first action i 0 but not on α 1 , and the last summation depends on both. In particular, the first row is the same as in (6) and (8), and thus the other terms correspond to the difference in the marginal costs between these actions.…”

Section: Propositionmentioning

confidence: 99%

“…As the dispatching decision does not depend on the state of the queues, SITA is a static (state-independent) policy and, assuming FCFS and Poisson arrivals, an expression for the mean waiting time follows from the Pollaczek-Khinchine mean value formula. SITA has been shown to be the optimal static size-aware policy for Poisson arrivals with respect to the mean waiting and sojourn time [11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Lookahead actions in dispatching to parallel queues

Hyytiä

2013

Performance Evaluation

View full text Add to dashboard Cite

Applying the first policy iteration (FPI) to any static dispatching (task assignment) policy yields a new improved dynamic policy that takes into account the particular cost structure and the expected future arrivals. However, it is generally hard to go beyond that due to the complex state space and the resulting difficulty in computing the value function for a dynamic policy. For example, applying FPI to identical FCFS servers with Bernoulli split gives the Least-Work-Left (LWL) policy, for which no closed-form value function is known. In fact, LWL with identical servers is equivalent to an M/G/k queue, the performance measures of which have remained as open problems. The situation gets even more complicated with heterogeneous servers. In this paper, we take an intermediate approach and consider lookahead actions that concern not only the current job but also the job arriving next, after which a basic (static) policy is assumed to take over. This is important as the benefits from some decisions can only be reaped with appropriate subsequent actions. The lookahead enables sound estimates also for marginal admission costs, e.g., with respect to LWL. The superior performance of the new near-optimal dispatching policies is demonstrated numerically.

show abstract

Section: Propositionmentioning

confidence: 99%

Section: Propositionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Lookahead actions in dispatching to parallel queues

Hyytiä

2013

Performance Evaluation

View full text Add to dashboard Cite

show abstract

“…Another important routing policy is Power of Two [10], [11], where for all incoming jobs, the dispatcher selects two servers independently and uniformly at random and applies the Join-the-Shortest-Queue policy among the chosen two servers. When the service demand is known and the servers are FCFS, the SITA policy with optimal thresholds is shown to optimize the performance of the system [12]. In this policy, each host serves jobs whose service demand is in a designated range.…”

Section: Related Workmentioning

confidence: 99%

Economies of scale in parallel-server systems

Doncel¹,

Aalto²,

Ayesta³

2017

IEEE INFOCOM 2017 - IEEE Conference on Computer Communications

View full text Add to dashboard Cite

Abstract-We consider a parallel-server system with K homogeneous servers where incoming tasks, arriving at rate , are dispatched by n dispatchers. Servers are FCFS queues and dispatchers implement a size-based policy such that the servers are equally loaded. We compare the performance of a system with n > 1 dispatchers and of a system with a single dispatcher. Every dispatcher handles a fraction 1/n of the incoming traffic and balances the load to K/n servers. We show that the performance of a system with n dispatchers, K servers and arrival rate coincides with that of a system with one dispatcher, K/n servers and arrival rate /n. Therefore, the performance comparison can be interpreted as the economies of scale in a system with one dispatcher when we scale up the number of servers and the arrival rate proportionately. We consider two continuous service time distributions: uniform and Bounded Pareto that have increasing and decreasing failure rates, respectively; and a discrete distribution with two values, which is the distribution that maximizes the variance for a given mean. We show that the performance degradation is small for uniformly distributed job sizes, but that for Bounded Pareto and two points distributions it can be unbounded.

show abstract

“…While a great many papers have been written comparing the response time of different task assignment policies, e.g., [5], [6], [9], [11], [12], [19], [22], [23], all of these papers conclude (via numerical analysis, simulation, or approximation) that, for high job-size variability, the SITA policy is superior to all the other common policies mentioned above. The reason for the superiority of SITA task assignment lies in the fact that SITA allows short jobs their own "expressline," thereby giving them isolation from long jobs.…”

Section: B the Advantages Of Sita For High Variabilitymentioning

confidence: 99%

Why segregating short jobs from long jobs under high variability is not always a win

Scheller‐Wolf

Young

2009

2009 47th Annual Allerton Conference on Communication, Control, and Computing (Allerton)

View full text Add to dashboard Cite

Abstract-This paper investigates the performance of task assignment policies for server farms as the variability of job sizes (service demands) approaches infinity. The Size-IntervalTask-Assignment policy (SITA), which separates short jobs from long jobs, has long been viewed as the panacea for dealing with high-variability job-size distributions. A very recent paper [16] showed that this common wisdom is flawed: SITA can actually be inferior to the much simpler greedy policy, Least-Work-Left (LWL), for certain common job-size distributions, including many modal, hyperexponential, and Pareto distributions.The above finding leads one to question whether providing isolation for short jobs from long ones is inherently bad, or whether it is just SITA's strict isolation of short jobs that sometimes leads to poor performance. To answer this question, we consider a much more flexible policy, which we call "CycleStealing" (CS). The CS policy is very similar to LWL, in that short jobs can go to any queue, but it still provides short jobs isolation from longs (one server is reserved for short jobs). While CS has many of the same properties as LWL, including high utilization of both servers, we prove, surprisingly, that, for high variability job sizes, CS performs poorly whenever SITA performs poorly. This result suggests that the notion of isolating short jobs from long jobs, under high variability workloads, is sometimes simply not the right thing to do.

show abstract

Optimal state-free, size-aware dispatching for heterogeneous -type systems

Cited by 76 publications

References 10 publications

Lookahead actions in dispatching to parallel queues

Lookahead actions in dispatching to parallel queues

Economies of scale in parallel-server systems

Why segregating short jobs from long jobs under high variability is not always a win

Contact Info

Product

Resources

About