Fedora 12 Scheduling Criteria & Algorithms

8/4/2019 Fedora 12 Scheduling Criteria & Algorithms

1/48

Click to edit Master subtitle style

Fedora 12 scheduling criteria& algorithms


2/48

4/21/12

Fedora 12 Specifications

Fedora 12 is a 0s based on theLinux kernel, developed by thecommunity-supported Fedora Projectand sponsored by Red Hat.

Fedora 12 uses Fedora Core 6development environment

Fedora core 6 uses the 2.6.18 basedLinux kernel
http://en.wikipedia.org/wiki/Linux_kernelhttp://en.wikipedia.org/wiki/Fedora_Projecthttp://en.wikipedia.org/wiki/Red_Hathttp://en.wikipedia.org/wiki/Red_Hathttp://en.wikipedia.org/wiki/Fedora_Projecthttp://en.wikipedia.org/wiki/Linux_kernel


3/48

4/21/12

History of Linux kernel

The Linux kernel was initiallyconceived and created by] Linus

Torvalds in 1991

He was influenced by the operating

system MINIX written by Andrew S.Tanenbaum
http://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Linus_Torvaldshttp://en.wikipedia.org/wiki/Linus_Torvalds


4/48

4/21/12

Linux kernel specification

Linux is a monolithic kernel

Preemptive kernel

Support symmetric multi processing

Written in C using GCC compiler
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


5/48

4/21/12

Linux kernel versions

The Linux Kernel Archives is the

official site to

Download standard Linux kernel.

3.0.1 is the latest stable kernelrealease

It was released in 05-08-2011

But we are taking about 2.6.37.6kernel in Fedora 12
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


6/48

4/21/12

We are going to talk moreabout scheduling Selecting processes one by one from

the ready queue and allocate CPU forthem.

The decisions may occur on

Process Switches from running towaiting state

Switches from running to ready state

Switches from waiting to ready Terminates
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


7/48

4/21/12

Scheduling Policy Process Preemption

When the dynamic priority of the currently runningprocess is lower than the process waiting in thereadyqueue

A process may also be preempted when its time quantumexpires

How Long Must a Time Quantum Last?

Quantum duration too short - system overhead caused bytask switching becomes excessively high

Quantum duration too long - processes no longer appearto be executed concurrently, degrades the response timeof interactive applications, and degrades theresponsiveness of the system

The rule of thumb adopted by Linux is: choose a duration

as long as possible, while keeping good system responsetime
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


8/48

4/21/12

What is Scheduling Criteria CPU utilization keep the CPU as

busy as possible Throughput Number of processesthat complete their execution per

time unit Turnaround time amount of timeto execute a particular process

Waiting time amount of time aprocess has been waiting in the readyqueue

Response time amount of time it
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


9/48

4/21/12

Scheduling Algorithems First Come First Serve Scheduling

Shortest Job First Scheduling

Priority Scheduling

Round-Robin Scheduling Multilevel Queue Scheduling

Multilevel Feedback-QueueScheduling
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


10/48

4/21/12

First Come First ServeScheduling (FCFS)Process Burst time

P1 24

P2 3

P2 3
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


11/48

4/21/12

First Come First ServeSchedulingThe average of waiting time in this

policy is usually quite long

Waiting time for P1=0; P2=24; P3=27 Average waiting time=(0+24+27)/3=17
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


12/48

4/21/12

First Come First ServeScheduling Suppose we change the order of

arriving job P2, P3, P1


13/48

4/21/12

First Come First ServeScheduling Consider if we have a CPU-bound process

and many I/O-bound processes

There is a convoy effect as all the other

processes waiting for one of the big processto get off the CPU

FCFS scheduling algorithm is non-preemptive
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


14/48

4/21/12

Short job first scheduling (SJF)This algorithm associates with each

process the length of the processesnext CPU burst

If there is a tie, FCFS is used In other words, this algorithm can bealso regard as shortest-next-cpu-burst

algorithm


15/48

4/21/12

Short job first scheduling SJF is optimal gives minimum

average waiting time for a given setof processes


16/48

4/21/12

ExampleProcesses Burst time

P1 6

P2 8

P3 7

P4 3

FCFS average waiting time: (0+6+14+21)/4=10.25

SJF average waiting time: (3+16+9+0)/4=7
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


17/48

4/21/12

Short job first scheduling

Two schemes:

Non-preemptive once CPU given to the

process it cannot be preempted untilcompletes its CPU burst

Preemptive if a new process arrives with

CPU burst length less than remaining timeof current executing process, preempt.This scheme is know as the Shortest-Remaining-Time-First (SRTF)


18/48


19/48

4/21/12

Short job first scheduling-Preemptive
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


20/48

4/21/12

Priority Scheduling

A priority number (integer) is associated witheach process

The CPU is allocated to the process with the

highest priority(smallest integer highest priority)

Preemptive

Non-preemptive

SJF is a special priority scheduling where priorityis the predicted next CPU burst time, so that itcan decide the priority


21/48

4/21/12

Priority SchedulingProcesses Burst time Priority Arrival time

P1 10 3

P2 1 1

P3 2 4

P4 1 5

P5 5 2

The average waitingtime=(6+0+16+18+1)/5=8.2
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


22/48

4/21/12

Priority SchedulingProcesses Burst time Priority Arrival time

P1 10 3 0.0

P2 1 1 1.0

P3 2 4 2.0

P4 1 5 3.0

P5 5 2 4.0

Gantt chart for both preemptive and non-preemptive, also waiting time


23/48

4/21/12

Priority Scheduling

Problem : Starvation low priority

processes may never execute

Solution : Aging as time progressesincrease the priority of the process


24/48

4/21/12

Round-Robin SchedulingThe Round-Robin is designed

especially for time sharing systems.

It is similar FCFS but add preemptionconcept

A small unit of time, called timequantum, is defined
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


25/48

4/21/12

Round-Robin Scheduling Each process gets a small unit of CPU

time (time quantum), usually 10-100milliseconds. After this time haselapsed, the process is preempted

and added to the end of the readyqueue.


26/48

4/21/12

Round-Robin Scheduling
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


27/48

4/21/12

Round-Robin Scheduling If there are n processes in the ready

queue and the time quantum is q,then each process gets 1/n of the CPUtime in chunks of at most q time units

at once. No process waits more than(n-1)q time units.
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


28/48

4/21/12


Performance

q large => FIFO

q small => q must be large withrespect to context switch, otherwise

overhead is too high

Typically, higher average turnaroundthan SJF, but better response


29/48

4/21/12

http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


30/48

Linux Version Scheduler

Linux pre 2.5 Multilevel Feedback Queue

Linux 2.5-2.6.23 O(1) Scheduler

Linux Post 2.6.23 Completely Fair Scheduler

Linux SchedulerHistoryLinux schedulers history is short but the popularityand its development achieved makes it impossible tosay that this was introduced in the later part of the19th century.

Below are some the schedulers' through its briefhistory
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


31/48

LINUX Schedulers

Linux operating system is a timesharing system thus the concept ofshort and long term schedulers istaken to higher level,

introducing an additional intermediary

level of scheduling or commonlyknown as medium term scheduling.

Key ideas behind medium term scheduling
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


32/48

Linux 1.2 Scheduler

Linux 1.2 scheduler used a circularqueue for runnable task managementthat operated with a round-robinscheduling policy.

.Features of 1.2 scheduler

Efficient for adding and removing processes (witha lock to protect the structure ).

Simple and fast .


33/48

2.2 and 2.4 Schedulers

Linux 2.2 scheduler

Linux 2.2 schedulers introduced the idea

of scheduling classes, permittingscheduling policies for real-time tasks,non-preemptive tasks, and non-real-timetasks.

Linux 2.4 scheduler


34/48

Linux 2.6 O(1) SchedulerReReducing scheduling algorithm

complexity to O(1)from O(n).Better support for SMP system.

Single run queue lock Cache problem

Preemptive: A higher priority processcan preempt a running process withlower priority.

P i i d i i i ff


35/48

Priority and interactivity effect ontime slice

Ni V l t ti i it
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


36/48

Nice Value vs. static priorityand Quantum

Static Priority NICE Quantum

High Priority 100 -20 800 ms

110 -10 600 ms

120 0 100 ms

120 +10 50 ms

Low Priority 139 +19 5 ms
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


37/48

4/21/12

Where Did We Come From?

Pre 2.6 Schedulers

Didnt utilize SMP very well

Single runqueue lock meant idleprocessors awaiting lock release

Preemption not possible

Lower priority task can execute whilehigh priority task waits

O(n) complexity

Slows down with larger input.


38/48

4/21/12

CFS The Future is Now! Completely Fair Scheduler

Merged into the 2.6.23 kernel

Runs task with the gravest need

Guarantees fairness (CPU usage)

No runqueues!

Uses a time-ordered red-black binarytree

Leftmost node is the next process to run


39/48

4/21/12

Red/Black Tree Rules

1

9

27

NL

NL

NL

NL

NL

NL

NL

NL

NL

98

49

2

3

4

65

31

25

7

NL
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


40/48

4/21/12

Red/Black Continues.

1. Every node has two children, each

colored either red or black.2. Every tree leaf node is colored

black.

3. Every red node has both of itschildren colored black.

4. Every path from root to a tree leaf

contains same number of blacknodes


41/48

4/21/12

CFS Features (cont) No timeslices!... sort of

Uses wait_runtime (individual) andfair_clock (queue-wide)

Processes build up CPU debt

Different priorities spend timedifferently

Half priority task sees time pass twice as

fast

O(log n) complexity

Only marginally slower than O(1) at very

large numbers of inputs

Digging in CFS Data
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


42/48

4/21/12

Digging in CFS DataStructures

CFS has three primary structures

task_struct, sched_entity, and cfs_rq. task_struct is the top-level entity,

containing things such as task

priorities, scheduling class, and thesched_entity struct. (sched.h, L1117)

sched_entity includes a node for the

RB tree and the vruntime statistic,among others. (sched.h, L1041)

cfs_rq contains the root node, task

group (more on this later), etc.


43/48

4/21/12

Priorities and more While CFS does not directly use

priorities or priority queues, it doesuse them to modulate vruntimebuildup.

In this version, priority is inverse toits effect a higher priority task willaccumulate vruntime more slowly,

since it needs more CPU time. Likewise, a low-priority task will have

its vruntime increase more quickly,

causing it to be preempted earlier.
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


44/48

4/21/12

...that's it?The CFS algorithm is, as stated, a lot

simpler than the previous one, anddoes not require many of the oldvariables.

Preemption time is variable,depending on priorities and actualrunning time. So we don't need

assign tasks a given timeslice.


45/48

4/21/12

Other additions CFS introduced group scheduling in

release 2.6.24, adding another levelof fairness.

Tasks can be grouped together, such

as by the user which owns them.CFS can then be applied to the grouplevel as well as the individual task

level. So, for three groups, it would give

each about a third of the CPU time,

and then divide that time up among
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


46/48

4/21/12

Modular scheduling Alongside the initial CFS release

came the notion of modularscheduling, and scheduling classes.

This allows various scheduling

policies to be implemented,independent of the genericscheduler.

sched.c, which we have seen,contains that generic code. Whenschedule() is called, it will callpick_next_task(), which will look atthe task's class and call the class-
http://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaumhttp://en.wikipedia.org/wiki/Andrew_S._Tanenbaum


47/48


48/48

Fedora 12 Scheduling Criteria & Algorithms

Documents

Transcript of Fedora 12 Scheduling Criteria & Algorithms