ch05_CPU_Scheduling.pdf

7/28/2019 ch05_CPU_Scheduling.pdf

1/29

Chapter 5: CPU SchedulingChapter 5: CPU Scheduling


2/29

Chapter 5: CPU SchedulingChapter 5: CPU Scheduling

Basic Concepts

Scheduling Criteria

Scheduling Algorithms

Multiple-Processor Scheduling

-

Thread Scheduling

Operating Systems Examples

Java Thread Scheduling

Algorithm Evaluation

5.2 Silberschatz, Galvin and Gagne 2005Operating System Concepts 7th Edition, Feb 2, 2005


3/29

Basic ConceptsBasic Concepts

Remember the benefit of multiprogramming?

z We have several processes in memory

z

When one blocks for I/O we can switch the CPU to anotherready process

Attem ts to maximize CPU utilization

Question: If we have many ready-to-run processes, which one to



4/29

Alternating Sequence of CPU And I/O BurstsAlternating Sequence of CPU And I/O Bursts

Process execution consists of a cycle of CPU execution and I/O wait

Initial CPU burst

Zero or more

CPU-I/O burst pairs

+



5/29

Histogram of CPUHistogram of CPU--burst Timesburst Times

For scheduling we will consider the length of the next CPU burst


Typically a few long bursts, many short ones


6/29

CPU SchedulerCPU Scheduler

Selects from among the processes in memory that are ready to

execute, and allocates the CPU to one of them

CPU scheduling decisions may take place when a process:

1. Switches from running to waiting state (e.g. system call, I/O)

2. Switches from runnin to read state e. . interru t

3. Switches from waiting to ready (e.g. I/O completes)

4. Terminatese per orm sc e u ng un er an

Under (2) and (3) we can:

z Allow the running process to continue non-preemptive

z Perform scheduling preemptive



7/29

CPU SchedulerCPU Scheduler

Non-preemptive means process continues until termination or I/O

Preemptive means that the process can be forced off the CPU

- ,

z Preemption could lead to complications: what if the process wasupdating a data shared with another process? E.g.,

u v , w

process P2

If P1 is preempted; s ar s execu ng

P2 accesses the shared variable, which is in inconsistent

state

Non-preemptive OS examples: Windows 3.x, MacOS before OS X

Preemptive OS examples: Windows 95 and later, Solaris, Linux



8/29

DispatcherDispatcher

Once a process is selected by the short-term scheduler, it is

(re)started by the dispatcher

z

saves state of currently running process (to its PCB)z loads state of selected process (from its PCB)

contextswitch

z switches to user mode

z jumps to the proper location in the selected process to restart itresume

The dispatch latency is time it takes for the dispatcher to stop one

process and start another running



9/29

Dispatch LatencyDispatch Latency



10/29

Scheduling CriteriaScheduling Criteria

What measures to use to select a scheduling algorithm?

CPU util ization % of time CPU is running user processes

To optimize

z On heavily-loaded computers want utilization 90%-100%

Throughput # of processes that complete their execution pertime unit

good

z depends on process length

z short processes: maybe 10 per second

Turnaround time lifespan of process

z Time from process submission to completion

z Includes execution time, I/O time, waiting on ready queue

Waiting time time spent on in the ready queue good

Response time time it takes from when a request wassubmitted until the first response is produced, not output (fortime-sharing environment)

good


z mpor an or n erac ve processes


11/29

FirstFirst--Come, FirstCome, First--Served (FCFS) SchedulingServed (FCFS) Scheduling

P1 24

P2 3

3

Suppose that the processes arrive at t=0 in the order: P1 , P2 , P3The Gantt Chart for the schedule is:

1 2 3

24 27 300

Process Burst Time Turnaround

Time

Waiting Time

P1 24 24 0

P2 3 27 24

P3 3 30 27


verage wa t ng t me: a t 1 + a t 2 + a t 3 =(0 + 24 + 27)/3 = 17

Average turnaround time: (24+ 27 + 30)/3 = 27


12/29

FCFS Scheduling (Cont.)FCFS Scheduling (Cont.)Suppose that the processes arrive in the order

P2 , P3 , P1

The Gantt chart for the schedule is:

P1P3P2

Waiting time forP1 = 6; P2 = 0; P3 = 3

63 300

urnaroun me or 1 = ; 2 = ; 3 =

Average waiting time: (6 + 0 + 3)/3 = 3

Average turnaround time: (30 + 3 + 6)/3 = 13

Much better than previous case

Convoy effect delays short process behind long process

z For example 1 CPU-bound, many I/O-bound processes


z Results in lower CPU and device utilization than letting shorter

processes execute first.


13/29

ShortestShortest--JobJob--First (SJF) SchedulingFirst (SJF) Scheduling

Associate with each process the length of its next CPU burst

Two versions:

z Non-preemptive once CPU given to the process it cannot be

preemp e un comp e es s urs

z preemptive if a new process arrives with CPU burst length

less than remaining time of current executing process, preempt Known as Shortest-Remaining-Time-First (SRTF)

SJF is optimal

z Gives minimum average waiting time for a given set of

processes (why?)



14/29

Example of NonExample of Non--Preemptive SJFPreemptive SJF

Process Arrival Time Burst Time

P 0.0 7

P2 2.0 4P3 4.0 1

4 .

SJF (non-preemptive)

1 3 2

73 160

4

8 12

Arrival Time should be considered -> Wait(Pi) = StartTime(Pi) - ArrivalTime(Pi)

Average waiting time = Wait(P1) + Wait (P2) + Wait (P3) + Wait (P4) / 4


= - - -

= (0 + 6 + 3 + 7) / 4 = 4


15/29

Example of Preemptive SJFExample of Preemptive SJF

Process Arrival Time Burst Time

P 0.0 7

P2 2.0 4P3 4.0 1

4 .

SJF (preemptive)

P1 P3P2

42 110

P4

5 7

P2 P1

16

Average waiting time = Wait(P1) + Wait(P2) + Wait(P3) + Wait(P4)

= (11 - 2) + (5 4) + (4 4) + (7 - 5)


= =


16/29

Determining Length of Next CPU BurstDetermining Length of Next CPU Burst

Can only estimate the length

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

exponential averaging

burstCPUoflengthactual1. = thn nt

10,3.

burstCPUnexttheforvaluepredicted2.

=+

1n

.1t +=:e ne. nnn=



17/29

Prediction of the Length of the Next CPU BurstPrediction of the Length of the Next CPU Burst



18/29

Examples of Exponential AveragingExamples of Exponential Averaging

=0

z n+1 = n

z Recent history does not count

=1

z + = t

z Only the actual last CPU burst counts

If we expand the formula, we get:

= t + 1 - t - 1 +

+(1 - )j tn -j +

+(1 - )n +1 0

Since both and (1 - ) are less than or equal to 1, eachsuccessive term has less weight than its predecessor



19/29

Priority SchedulingPriority Scheduling

A priority number is associated with each process

Process with the highest priority is selected

Priority: usually integer

In UNIX and textbook, smallest integer highest priority

Can be:

z Preemptive

z Non-preemptive

SJF is a s ecial case of riorit schedulin . Wh ?

z Priority is the predicted next CPU burst time

z Larger the CPU burst the lower the priority

z It is possible that a low priority process never executes

Solution: aging


,


20/29

Round Robin (RR)Round Robin (RR)

Designed for time-sharing systems

Each process gets a small unit of CPU time q (time quantum),

- .

If quantum expires: preempt process, add to end of ready queue If process executes less than q: reschedule

Ready queue is FIFO

If there are n processes in the ready queue, each process gets

1/n of the CPU time in chunks of at most q time units at once.o process wa s more an n- q me un s.

Ifq= then RR becomes what scheduling algorithm?

z q FCFS

z q small q must be large with respect to context switch,otherwise overhead is too high



21/29

Example of RR with Time Quantum = 20Example of RR with Time Quantum = 20

Process Burst Time

1

P2 17

P3 68

P4 24

The Gantt chart is:

P1 P2 P3 P4 P1 P3 P4 P1 P3 P3

0 20 37 57 77 97 117 121 134 154 162

Typically, higher average turnaround than SJF, but betterresponse



22/29

Time Quantum and Context Switch TimeTime Quantum and Context Switch Time

Smaller time quantum increases context switches.



23/29

Turnaround Time Varies With The Time QuantumTurnaround Time Varies With The Time Quantum

Turnaround time varies with the time quantum.



24/29

Multi level QueueMulti level Queue

Ready queue is partitioned into separate queues:

foreground (interactive)

ac groun a c

Each queue has its own scheduling algorithm

z foreground RR

z background FCFS

Scheduling must be done between the queues

. .,

from background). Possibility of starvation.

z Time slice each queue gets a certain amount of CPU time

which it can schedule amon st its rocesses; i.e., 80% to

foreground in RR

z 20% to background in FCFS



25/29

Multilevel Queue SchedulingMultilevel Queue Scheduling



26/29

Multi level Feedback QueueMulti level Feedback Queue

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

Multilevel-feedback-queue scheduler defined by the followingparameters:

z scheduling algorithms for each queue

z method used to determine when to upgrade a processz method used to determine when to demote a process

z method used to determine which queue a process will enter

when that process needs service



27/29

Example of Multi level Feedback QueueExample of Multi level Feedback Queue

Three queues:

z Q RR with time quantum 8 milliseconds

z Q1 RR time quantum 16 milliseconds

z Q2 FCFS

z A new job enters queue Q0 which is served FCFS. When it

gains CPU, job receives 8 milliseconds. If it does not finish in 8, .

z At Q1job is again served FCFS and receives 16 additional

milliseconds. If it still does not complete, it is preempted and

moved to ueue Q .



28/29

Multilevel Feedback QueuesMultilevel Feedback Queues



29/29

End of Chapter 5End of Chapter 5

Documents

ch05_CPU_Scheduling.pdf