Talk:Scheduling algorithm

Work/non-work conserving schedulers in communication systems

In wireless communication systems, scheduling is an important concept. The basic task of a Scheduler is to allocate the radio resource among users. The function of scheduler in wireless networks is two fold: First to provide certain minimum required Quality of Service (QoS) to the user and second to maximize the system throughput. As wireless networks work in a traditional layered model, the scheduler works at medium access control (MAC) layer. There are many scheduling algorithms in the literature but mainly wireless schedulers can be classified in two broad categories.

1. Work - Conserving Schedulers. 2. Non - Work Conserving Schedulers.

A work conserving scheduler never remains idle if there is a packet to be transmitted in the queue. A non - work conserving scheduler is one which remains idle even if there is a packet to be transmitted in the queue while scheduler may be expecting for another packet of high priority.

137.111.13.34 05:28, 18 June 2006 (UTC)

Okay. Non-work conserving (or conservative?) scheduling are only used in circuit-oriented TDMA systems, right? Any good examples? GSM?

I suppose that work conserving scheduling is used in packet-switching, not only best-effort, but also systems where resources are reserved in advance (for example GPRS and ATM), and systems that offer guaranteed quality of service, since remaining capacity can be given to users that do not require any special service.

Mange01 07:45, 21 October 2006 (UTC)

Separate communication article?

I suggest that the data packet scheduling issues are removed from scheduling (computing) and scheduling algorithm articles into a separate article on scheduling (communication). Perhaps the remaining scheduling algorithm article should be merged with [[[scheduling (computing)]], which now should focus on operational systems and multitasking.

Which of all the scheduling disciplines, policies and algorithms listed in the scheduling (computing) article belongs to networking, and which belongs to operational systems?

Is it possible to clarify what is an algorithm (such as fair queuing), and what is a policy or discipline (such as max-min fairness)?

Mange01 07:25, 21 October 2006 (UTC)

I would like to catch your attention to one of the key area of scheduling which we can not ignore, the scheduling of time for institutes and organizations. A lot of projects pending because of Non-solvability of scheduling problem, like one of the major IT firm is unable to schedule the time table for MIT.

I'm removing the split tag and editing the article to generalize it for communications and computing. Hopefully someone will come by and instantiate the Scheduling (communications) redlink for you. --Kvng (talk) 22:06, 2 August 2008 (UTC)

Stub

This article section:

http://en.wiki.x.io/wiki/CPU_scheduler#Scheduling_disciplines

Places this article as it's main article for a section, but this article is smaller than the section.

I don't know enough of the subject to know if the content there belongs here... could somebody who knows move it over?--DustWolf (talk) 15:25, 29 June 2008 (UTC)

Example CPU allocation

So... if an application creates three threads -- one GUI thread and two background threads handling some long process -- how does the scheduler allocate two CPUs for these three threads?

Does it run a thread on a CPU for the thread's lifetime or can it resume a thread's execution on a different CPU during a context switch? Could the GUI thread wind up with an entire CPU to itself while the two background threads are forced to share the other CPU? —Preceding unsigned comment added by 96.48.16.201 (talk) 23:58, 14 October 2009 (UTC)

Back to basics

Why does the article state that in computer science, scheduling algorithms are "for" allocating resources to data flows and similar constructs? This is simply not true. Computer resource allocation is merely one of the applications of scheduling, which include things like: scheduling the movements of materials, as in logistics; scheduling classes with regards to constraints such as teacher/room availability and school hours; and many other similar problems. The algorithms involved in these problems' solutions are just as much a part of "computer science" as the ones presented. — Preceding unsigned comment added by 75.145.77.185 (talk) 21:23, 6 July 2011 (UTC)