Chapter 3: Process

Scheduling

Operating System Concepts – 9th Silberschatz, Galvin and Gagne ©2013

Edition

Chapter 3: Process Scheduling

n Basic Concepts
n Scheduling Criteria
n Scheduling Algorithms
n Multiple-Processor Scheduling
n Real-Time CPU Scheduling

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Basic Concepts
process
n Maximum CPU utilization obtained
the H
with multiprogramming
- E PressT
n CPU–I/O Burst Cycle – Process
execution consists of a cycle of CPU
execution and I/O wait

n 載入中⋯
CPU burst followed by I/O burst
n CPU burst distribution is of main
concern

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
這張幻燈片講解了有關 CPU 和 I/O 營運中的⼀些基本概念，特別是CPU-I/O 突發循環（CPU-I/O Burst Cycle）。讓我⼀步步解釋：

1. 多程操作下的最⼤ CPU 利⽤率

• 當使⽤多程（Multiprogramming）時，系統可以同時運⾏多個進程，從⽽實現 CPU 的最⼤利⽤率。多程操作允許 CPU 在⼀個進程

等待 I/O 操作時運⾏另⼀個進程，從⽽減少閒置時間。

2. CPU-I/O 突發循環（CPU-I/O Burst Cycle）

• 進程的執⾏由CPU 突發（CPU burst）和I/O 突發（I/O burst）交替組成。每當進程使⽤ CPU 時，這就是 CPU 突發；當進程在等待

I/O 操作（例如讀寫⽂件）時，這就是 I/O 突發。

• 這形成了⼀個循環（cycle）：進程在 CPU 上運⾏（CPU 突發），然後進入 I/O 操作階段（I/O 突發），如此往復。

3. CPU 突發和 I/O 突發的順序

• 典型的進程會有⼀個CPU 突發，隨後進入⼀個I/O 突發階段，然後可能再次進入 CPU 突發。

• 圖片右邊展⽰了這個順序：

• 進程在 CPU 上執⾏加載、存儲、加法等操作（CPU 突發），然後等待 I/O 操作（I/O 突發，如讀⽂件）。

• 當 I/O 完成後，進程再次在 CPU 上執⾏（例如增加索引、寫入⽂件等），隨後再⼀次進入 I/O 等待。

• 這個過程⼀直循環，直到進程完成所有操作。

4. CPU 突發的分佈是主要關注點

• CPU 突發分佈是進程調度的主要關注點。不同的進程有不同的 CPU 和 I/O 突發長度，調度程序需要根據這些特性來做出最佳調度決

策，以保持 CPU 的⾼效利⽤。

總結：

CPU 和 I/O 突發的交替循環是進程執⾏的基本特性。通過多程操作，操作系統

可以在⼀個進程等待 I/O 時運⾏另⼀個進程，從⽽最⼤化 CPU 利⽤率。進程的
調度策略基於這些突發週期，以最佳⽅式管理 CPU 資源。

如果你有其他問題或想了解更多細節，請告訴我！
Edition

Histogram of CPU-burst Times

IO bound
Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

CPU Scheduler
n Short-term scheduler selects from among the processes in ready queue, and
allocates the CPU to one of them
l Queue may be ordered in various ways
n CPU scheduling decisions may take place when a process:
1. 12 Switches from running to waiting state

2. 2 Switches from running to ready state

-

3. X Switches from waiting to ready

n
4. ↑ Terminates
r

載入中⋯
Scheduling under 1 and 4 is nonpreemptive EFEEt
n All other scheduling is preemptive -ItG'7
l Consider access to shared data
l Consider preemption while in kernel mode
l Consider interrupts occurring during crucial OS activities

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Diagram of Process State

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Representation of Process Scheduling

n Queuing diagram represents queues, resources, flows

↓ CPU
scheduler

RR

Faiting time
Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

15
Dispatcher
E process BE BEL DUEDX
pu schedule
E process
n Dispatcher module gives control of the CPU to the process selected by the
short-term scheduler; this involves:
l switching context
l switching to user mode
l jumping to the proper location in the user program to restart that program

n Dispatch latency – time it takes for the dispatcher to stop one process and
start another running

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Scheduling Criteria
# * G
n CPU utilization – keep the CPU as busy as possible
number of
E
n Throughput – # of processes that complete their execution per
time unit
⑰
n Turnaround time – amount of time to execute a particular process
⑰
n Waiting time – amount of time a process has been waiting in the
ready queue
⑭
n Response time – amount of time it takes from when a request was
submitted until the first response is produced, not output (for time-
sharing environment)

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Scheduling Algorithm Optimization Criteria

n Max CPU utilization

n Max throughput
n Min turnaround time
n Min waiting time
n Min response time

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

First-Come, First-Served (FCFS) Scheduling

A
Process Burst Time
P1 24
P2 3
P3 3

n Suppose that the processes arrive in the order: P1 , P2 , P3

The Gantt Chart for the schedule is:

載入中⋯
P1 P2 P3

0 24 27 30
n Waiting time for P1 = 0; P2 = 24; P3 = 27
n Average waiting time: (0 + 24 + 27)/3 = 17

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

FCFS Scheduling (Cont.)

Suppose that the processes arrive in the order:
P2 , P3 , P1
n The Gantt chart for the schedule is:

P2 P3 P1

0 3 6 30

n Waiting time for P1 = 6; P2 = 0; P3 = 3

n Average waiting time: (6 + 0 + 3)/3 = 3
n Much better than previous case
n Convoy effect - short process behind long process
l Consider one CPU-bound and many I/O-bound processes

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Shortest-Job-First (SJF) Scheduling

TEY
n Associate with each process the length of its next CPU burst
l Use these lengths to schedule the process with the shortest time

n SJF is optimal – gives minimum average waiting time for a given set of
processes
l The difficulty is knowing the length of the next CPU request
l Could ask the user

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Example of SJF
ProcessArriva l Time Burst Time
P1 0.0 6
P2 2.0 8
P3 4.0 7
P4 5.0 3
n SJF scheduling chart

P4 P1 P3 P2

0 3 9 16 24

n Average waiting time = (3 + 16 + 9 + 0) / 4 = 7

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Determining Length of Next CPU Burst

n Can only estimate the length – should be similar to the previous one
l Then pick process with shortest predicted next CPU burst

n Can be done by using the length of previous CPU bursts, using exponential averaging

1. t n = actual length of n th CPU burst

2. τ n +1 = predicted value for the next CPU burst
3. α , 0 ≤ α ≤ 1 T#2tEAGETE
4. Define :
τ n +1 = α t + (1 − α )τ .
n n

n Commonly, α set to ½

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Prediction of the Length of the

Next CPU Burst

to to

i
Given =10 to Ez "

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Examples of Exponential Averaging

n & =0
l Tn+1 = Tn
l Recent history does not count
n & =1
l T n+1 = d tn

l Only the actual last CPU burst counts

n If we expand the formula, we get:
-(d)
n+1 = tn+(1 - ) tn -1 + …
+(1 - & )j d)
*

tn -j + …
+(1 - & )n +1 M 0
n Since both & and (1 - d ) are less than or equal to 1, each successive term has
less weight than its predecessor
n Preemptive version called shortest-remaining-time-first
FARSJF
shortest job first

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Example of Shortest-remaining-time-first
-

n Now we add the concepts of varying arrival times and preemption to the analysis

ProcessA arri Arrival TimeT Burst Time

II
P1 0 8
P2 1 4 1 .
4
P3 2 9 3 ,
7
,
9
P4 3 5 2
, 5,
9
,
9 + 2 = 26
n Preemptive SJF Gantt Chart

P1 P2 P4 P1 P3

0 1 5 10 17 26
n Average waiting time = [(10-1)+(1-1)+(17-2)+(5-3)]/4 = 26/4 = 6.5 msec

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Priority Scheduling
n
can be preentive
A priority number (integer) is associated with each process
or
impreentive
n The CPU is allocated to the process with the highest priority (smallest integer
highest priority)
l Preemptive
l Nonpreemptive

n SJF is priority scheduling where priority is the inverse of predicted next CPU
burst time

n Problem Starvation – low priority processes may never execute

n Solution Aging – as time progresses increase the priority of the process

=ZEE - > priority it My
process

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Example of Priority Scheduling

ProcessA arri Burst TimeT Priority
P1 10 3
P2 1 1
P3 2 4
P4 1 5
P5 5 2
n Priority scheduling Gantt Chart

P2 P5 P1 P3 P4

n Average
0 1waiting time
6 = 8.2 msec 16 18 19

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Round Robin (RR)

n Each process gets a small unit of CPU time (time quantum q), usually 10-100
milliseconds. After this time has elapsed, the process is preempted and added
to the end of the ready queue.
n If there are n processes in the ready queue and the time quantum is q, then
each process gets 1/n of the CPU time in chunks of at most q time units at
once. No process waits more than (n-1)q time units.
n Timer interrupts every quantum to schedule next process
n Performance
l q large FCFS
l q small q must be large with respect to context switch, otherwise
overhead is too high
context switch
ENTE

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Example of RR with Time Quantum = 4

Process Burst Time

P1 24
P2 3
P3 3

n The Gantt chart is:

P1 P2 P3 P1 P1 P1 P1 P1

n 0
Typically, 4
higher 7
average 10 14than SJF,
turnaround 18 but22 better26 30
response
n q should be large compared to context switch time
n q usually 10ms to 100ms, context switch < 10 μsec

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Time Quantum and Context Switch Time

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Turnaround Time Varies With

The Time Quantum

l q large FIFO
l q small q must be large with respect to context
switch, otherwise overhead is too high
l 80% of CPU bursts should be shorter than q
1 -

FLE
2 .
354
3 . SRTF
4 :
Priority
5 .
RR

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
 Co
11 13 of 19

FEEL
(a) jox/05 is

er tt
SJF

Priority
Edition

Multilevel Queue
n Ready queue is partitioned into separate queues, eg:
l foreground (interactive)
l background (batch)
n Process permanently in a given queue

n Each queue has its own scheduling algorithm:

l
l
foreground – RR
background – FCFS
quene m j scheduling
quenezr
-

queye A) scheduling
n
l
-
Scheduling must be done between the queues:
I Fixed priority scheduling; (i.e., serve all from foreground then from
background). Possibility of starvation.
l ZaTime slice – each queue gets a certain amount of CPU time which it can
schedule amongst its processes; i.e., 80% to foreground in RR
l 20% to background in FCFS

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Multilevel Queue Scheduling

foreground

background

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
 priority
high system
processes
-

I
interactive processes
internative editing processes
butch procesty
student
low
process
Edition

Multilevel Feedback Queue

n A process can move between the various queues; aging can be

implemented this way

n Multilevel-feedback-queue scheduler defined by the following parameters:

l (e number of queues
l 2 scheduling algorithms for each queue
-

l 3- method used to determine when to upgrade a process Equeni

l - method used to determine when to demote a process
4

l 5- method used to determine which queue a process will enter when that

process needs service

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Example of Multilevel Feedback Queue

n Three queues:
l Q0 – RR with time quantum 8 milliseconds
l Q1 – RR time quantum 16 milliseconds
l Q2 – FCFS

n Scheduling
l A new job enters queue Q0 which is
served RR
4 When it gains CPU, job receives 8
milliseconds
4 If it does not finish in 8 milliseconds,
job is moved to queue Q1
l At Q1 job is again served RR and receives
16 additional milliseconds
4 If it still does not complete, it is -
preempted and moved to queue Q2

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Real-Time CPU Scheduling

n Can present obvious challenges
n Soft real-time systems – no
guarantee as to when critical real-
time process will be scheduled
n Hard real-time systems – task
must be serviced by its deadline
n Two types of latencies affect
performance E 載入中⋯
1. Interrupt latency – time from #HA
arrival of interrupt to start of
routine that services interrupt Interrupt Service Routine

2. Dispatch latency – time for

schedule to take current process
off CPU and switch to another

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

E Real time process E. E BERE Set real time syste

Priority-based Scheduling
-

n For real-time scheduling, scheduler must support preemptive, priority-based

scheduling F
l But only guarantees soft real-time
n For hard real-time must also provide ability to meet deadlines
n Processes have new characteristics: periodic ones require CPU at constant
intervals
l Has processing time t, deadline d (optional), period p
l 0≤t≤d≤p
l Rate of periodic task is 1/p
i

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Rate Monotonic Scheduling

-
f(x) =
Y
n A priority is assigned based on the inverse of its period

n ~ Shorter periods = higher priority; 67 79

n Longer periods = lower priority

n Example 1
Given two processes P1 and P2, if p1 = 50, p2 = 100 and t1 = 20, t2 =35,
Supposed priority of P2 is higher than P1,
Is it work? Ans: it’s not work!!

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Rate Monotonic Scheduling

su
Given two processes P1 and P2, if p1 = 50, p2 = 100 and t1 = 20, t2 =35,

if P1 is assigned a higher priority than P2.

Calculate the CPU utilization (ti/pi), P1 is 20/50 = 0.4, P2 is 35/100 = 0.35

Total is 0.4 + 0.35 = 0.75 = 75% < 100% > IY TE

We could use Rate Montonic Scheduling, i.e., P1 has higher priority than P2
P p P

↑ i ↑
P . P PE Pipz4
EPIAZdenMineE EP PLAZ deadline
.

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
+
=

It
8,
o
·
↑

Pl D2 Pl
D)
↓ ↓
P2
Pl
P2
&
tso
20 to
50 do 90 80 90 10010 DO Br40150160190180190200 10 220230
Edition

Missed Deadlines with

Rate Monotonic Scheduling
n Example 2
Given two processes P1 and P2,
if p1 = 50, p2 = 80 and t1 = 25, t2 =35,

Calculate the CPU utilization (ti/pi), P1 is 25/50 = 0.5, P2 is 35/80 = 0.44

Total is 0.5 + 0.44 = 0.94 = 94% < 100%
Could use Rate Montonic Scheduling? ANS: NO!!!

27
↑ ↑ -10 /

25

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Rate Montonic Scheduling

n Rate montonic scheduling may not maximize the CPU
utilization
n Some restrictions

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Earliest Deadline First Scheduling (EDF)

n Priorities are assigned according to deadlines:

the earlier the deadline, the higher the priority;

the later the deadline, the lower the priority

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Earliest Deadline First Scheduling (EDF)

n Example 2
Given two processes P1 and P2, ce
if p1 = 50, p2 = 80 and t1 = 25, t2 =35, v

process #z deadline isCheck p1

In RMS, P1 has higher priority, we will and p2
choose P1!!

ppedendline
M ↑ my
,F
p Pict deadline FREE
However, in EDF, we check remaining deadlines, P2
↑ EDI J FidelineEH YETED2tdeadline MES BHIP , Fr
has higher priority,
(p2=80 and p1=100, we will choose P2!!)

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Proportional Share Scheduling

n T shares are allocated among all processes in the system

n An application receives N shares where N < T

n This ensures each application will receive N / T of the total processor time

n Example
l Suppose T= 100 shares, given three processes A, B, C
l We can allocate A 50 (50% CPU total processor time), B 15 (15% CPU total
processor time), and C 20 (20% CPU total processor time)
l For T= 100 shares, we have allocated 85 shares
l If a new process D needs 30 shares, OS will reject

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013
Edition

Proportional Share Scheduling

n T shares are allocated among all processes in the system

n An application receives N shares where N < T

n This ensures each application will receive N / T of the total processor time

n Example
l Suppose T= 100 shares, given three processes A, B, C
l We can allocate A 50 (50% CPU total processor time), B 15 (15% CPU total
processor time), and C 20 (20% CPU total processor time)
l For T= 100 shares, we have allocated 85 shares
l If a new process D needs 30 shares, OS will reject

Operating System Concepts – 9th 6.0 Silberschatz, Galvin and Gagne ©2013