236605 Advanced Topics Spring Semester, 2000 1.1 Shortest Remaining Processing Time (srpt) Rule
نویسندگان
چکیده
1 Algorithms for a Single Machine We show that simple greedy algorithms are optimal for some scheduling problems on a single machine with release dates and preemptions. SRPT Rule: At each moment, process the job with the shortest remaining processing time. Note that the implementation of SRPT is possible only when preemptions are allowed. Theorem 1.1 ((1]) SRPT rule is optimal for 1jr j ; pmtnj P j C j. Proof: Consider a schedule, S 1 , which is not according to SRPT. Assume that on time t, a job k with remaining processing time p 0 k is processed, while some other available job, j, with shorter remaining processing time, p 0 j , is waiting. Consider all the time intervals from t and ahead, in which j or k are processed. We reallocate these intervals to j and k in the following way: rst allocate intervals to j until it is completed, then allocate the remaining intervals to k. Since both jobs are completed in the schedule S 1 , the total length of these intervals is p 0 k + p 0 j , therefore, both jobs are completed also in the new schedule. We do not change the schedule of jobs other than j; k, thus, Fact 1.2 For any job i 2 J n fj; kg, the completion time of i is not aaected from the reallocation of the intervals. Claim 1.3 The value of C j + C k is reduced. Proof: Let c be the completion time of j; k in the schedule S 1. In the new schedule, the completion time of k is C k = c. Since p 0 j < p 0 k and in the schedule S 1 the job k is processed at time t, the completion time of the rst job out of j; k in S 1 is larger than the completion time of j in the new schedule (which is now the completion time of intervals of total length p 0 j). The reallocation of the intervals follows the SRPT rule. We conclude that any schedule that does not follow the SRPT rule can be converted into a one that follows SRPT in which P j C j is smaller. Note that we need to check which job achieves the SRPT only when the processing of the current job is completed or when a …
منابع مشابه
236605 Advanced Topics Spring Semester, 2000 2 the Approximation Scheme 2.1 Approximation for a Constant Range of Ratios Claim 2.1 Let S Be a Schedule for Which W J
1 Background In the previous lecture we saw a PTAS for the problem PkM 2 i. Recall that M i is the completion time of machine i. In this lecture we present a PTAS for the problem Pk P c j w j , due to Skutella and Woeginger (1999). Some known results for related problems: 1k P w j c j-can be solved in polynomial time using Smith's Ratio Rule. Pk P c j-can be solved in polynomial time using SPT ...
متن کاملImplementation of SRPT Scheduling in Web Servers
This paper proposes a method for improving the performance of Web servers servicing static HTTP requests. The idea is to give preference to those requests which are quick, or have small remaining processing requirements, in accordance with the SRPT (Shortest-Remaining-Processing-Time) scheduling policy. The implementation is at the kernel level and involves controlling the order in which socket...
متن کاملA large-deviations analysis of the GI/GI/1 SRPT queue
We consider a GI/GI/1 queue with the shortest remaining processing time discipline (SRPT) and light-tailed service times. Our interest is focused on the tail behavior of the sojourn-time distribution. We obtain a general expression for its large-deviations decay rate. The value of this decay rate critically depends on whether there is mass in the endpoint of the service-time distribution or not...
متن کاملSRPT applied to bandwidth-sharing networks
We consider bandwidth-sharing networks, and show how the SRPT (Shortest Remaining Processing Time) discipline can be used in order to improve the delay performance of the system. Our main idea is not to use SRPT globally between the traffic classes, which has been shown to induce instability, but rather deploy SRPT only locally within each traffic class. We show that with this approach, the per...
متن کاملDiffusion Limits for Shortest Remaining Processing Time Queues
We present a heavy traffic analysis for a single server queue with renewal arrivals and generally distributed i.i.d. service times, in which the server employs the Shortest Remaining Processing Time (SRPT) policy. Under typical heavy traffic assumptions, we prove a diffusion limit theorem for a measure-valued state descriptor, from which we conclude a similar theorem for the queue length proces...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2007